You are on page 1of 44

The Singing of Psalms in the Early-MedievalOffice

By Joseph Dyer The anonymous fourth-century tract De poenitentia depicts with a grand oratorical flourish the universality of the psalms in the Christian world. Its author proclaims that the psalms are heard at vigils, at morning prayer, and at funerals in the city churches, but the power of "David" extends beyond their walls. Even in the fields and deserts and stretchinginto uninhabitedwasteland,he rouses sacred choirs to God. ... In the monasteriesthere is a holy chorus of angelic hosts, and David is first and middle and last. In the convents . . David is first and middle and last. In the deserts . . . David is first and middle and last. And at night all men are dominated by physicalsleep and drawn into the depths, and David alone stands by, arousing all the servantsof God to angelic vigils, turning earth into heaven and making angels of men.' Though the author may have exaggerated the centrality of the psalms in the lives of all Christians, he aptly characterized the role of the psalms in monastic life: there David was indeed "first and middle and last." Every monk was expected to memorize all 150 psalms. They were his daily bread, words always on his lips, the foundation of his life of prayer. St. Benedict, taking as his model the monasteries attached to the Roman basilicas, recommended the weekly recitation of the psalter, and some monastic regimens were even more demanding. The psalms also held a place of fundamental importance in education. If a candidate for the monastic or clerical state could not read, the psalter served as his primer of Latin grammar, and facility in memorizing the psalms provided the clearest early indication of exceptional intellectual ability. In the medieval Office the singing of psalms was far more than a musical exercise. It was ampler in its connotations than the mere adaptation of words to stereotypical melodic formulae. Years of daily encounters with the prayers of the psalmist fostered a rich contextuality of associations, a private and
This essay is an offering to Janet Knapp from one of her former students on the occasion of her retirement from the faculty of Vassar College. Research support has been received from the Summer Stipend program of the National Endowment for the Humanities and from the University of Massachusetts. I would also like to acknowledge the assistance of Professor Ruth Steiner, who facilitated access to the Dom Mocquereau Microfilm Archive at the Catholic University of America. De poenitentia, PG 64:12-13. The translation is by James McKinnon, Music in Early Christian Literature,Cambridge Readings in the Literature of Music (Cambridge, Eng., 1987), p. 90. The number of patristic references to the psalms in this volume is itself revealing.

64 SPECULUM (1989)


The Singing of Psalms interior exegesis of the scriptural text in an ever-widening field of significance. The fruits of this meditation on the psalms rarely appeared in written form, but it resonated with the method of exegesis in the writings of the Fathers. The Fathers regarded the psalter as a book of prophecy, "an assortment of oracles whose meaning is revealed to those who have the insight to discern it and apply it to the contemporary scene."2 This prophetic and spiritual mode of interpretation received its sanction directly from the New Testament. Jesus applied the words of the psalmist to himself, and the early church saw the entire Old Testament as a repository of types which came to fulfillment with the mission of the Savior. Of the four forms of scriptural exegesis cultivated in the patristic age (historical, allegorical, moral, and anagogical), the allegorical and moral appealed most strongly to the monastic imagination.3 They encouraged richly allusive interpretations which led far from the historical meaning of the text. The psalms lent themselves especially well to this approach, and those who sang the psalter weekly in the Divine Office could refer to written models as they strove to apply the psalm text to the understanding of doctrine, the practice of virtue, or the life of the church. A range of patristic commentaries on the psalms was available in Latin to western monks of the Middle Ages. Origen was the first to comment on the entire psalter, and his method of drawing out the hidden significance of the text became the paradigm for all later commentators. Though he wrote in Greek, the essentials of his method were available in the West through Jerome's Tractatusin psalmos, an adaptation of Origen's homilies on the psalms, and in the Latin translation of the homilies on Psalms 36, 37, and 38 by Rufinus.4 Origen deeply influenced some of the most widely read of earlymedieval authors: Gregory the Great, Isidore, and Hrabanus Maurus. As Jean Leclercq observed, "what was sought in him was not so much a doctrine as a mentality, and, most of all, a way of interpreting Holy Scripture."5 This can be illustrated by examining briefly the beginning of Origen's homily on Psalm 36. After describing the three types of interpretation he recognized - proG. W. H. Lampe, "The Exposition and Exegesis of Scripture to Gregory the Great," Cambridge History of the Bible, 2 (Cambridge, Eng., 1969), p. 159. See also Joseph Gelineau, "Les psaumes a l'epoque patristique," La Maison-Dieu 135 (1978), 99-116; Balthasar Fischer, "Le Christ dans les psaumes: La devotion aux psaumes dans l'eglise des martyrs," La Maison-Dieu 28 (1951), 86109; and Pierre Salmon, L'Office divin au moyenage, Lex orandi 27 (Paris, 1959), pp. 99-134. 3 Henri de Lubac, Exegese medievale:Les quatre sens de l'Ecriture,2 vols. (Paris, 1959); Johannes Quasten, Patrology, 3 vols. (1950; repr. Utrecht, 1964), 2:92-93. 4Jerome, Tractatus in psalmos, CCSL 78 (Turnhout, 1958); Rufinus, Origenis explanatio (PG 12:1319-1410). Rufinus's Prologus is printed separately in CCSL 20 (Turnhout, 1961). See Vittorio Peri, Omelie origeniane sui salmi: Contributoall'identificazionedel testo latino, Studi e Testi 289 (Vatican City, 1980), pp. 7-40. The basic study of the patristic commentaries is MarieJoseph Rondeau, Les commentairespatristiques du Psautier (3e-5e siecles), Orientalia Christiana analecta 219-20 (Rome, 1982-85); a briefer overview may be found in Aime Solignac, "Psaumes, commentaires," Dictionnaire de spiritualite, 12/2:2562-68. 5 Jean Leclercq, The Love of Learning and the Desirefor God: A Study of Monastic Culture, trans. Catherine Misrahi (New York, 1961), p. 96; de Lubac, Exegese medievale, 1:221-38.


The Singing of Psalms 537 phetic, mystic, and moral - Origen said of Psalm 36: "this entire psalm is moral, and given to the human soul as a cleansing and a remedy, since it makes manifest our sins and teaches us to live in accordance with the law."6 The opening verses of this psalm are: "Do not strive to outdo the evildoers or emulate those who do wrong. For like grass they soon wither, and fade like the green of spring." The psalmist's comparison suggested to Origen a passage from Isaiah 40.6-8 ("All mankind is grass .... The grass withers, the flowers fade, but the word of our God endures for evermore"). The flower's bloom represents the "glory of the flesh" exemplified in the arrogance of princes. One reigns after another, but their "bloom" soon shrivels, turns to dust, and is scattered by the wind ("verum etiam tanquam pulvis aridus et a vento dispersus"). The rich and the vain enjoy themselves in this life, though soon enough even the location of their tombs is forgotten. Origen contrasted this evanescence with the stability of the word of God. Dumb animals feed on the grass but the wise man feeds on the divine word, which is eternal, and on Jesus, the bread which came down from heaven. Origen's exposition of this psalm illustrates a kind of exegesis which was widely admired and emulated. As Dom Leclercq explains, "if he [Origen] was the favorite model of monastic commentators, this was because of his mastery of allegory, and consequently of the whole theory of the spiritual life."7 Among the Fathers the psalter was the book most often commented upon, and the wealth of patristic reflection was transmitted to the Middle Ages both directly and through excerpts in later authors. Hilary of Poitiers probably treated the entire psalter, and considerable portions of his Tractatus are preserved. Ambrose commented on several psalms, and Jerome was known to medieval readers both for his Commentarioliand for the Breviarium in psalmos once attributed to him. Augustine's prestige ensured the dissemination of his Enarrationes, the first completely preserved Latin commentary on the entire psalter. His writings gained even wider currency when Cassiodorus borrowed from them for his own influential exegesis of the psalms. Augustine, Jerome, and Cassiodorus became the source books for Carolingian commentators on the psalms.8 All monks would have had access to at least some of these commentaries or their literary descendants. Anyone who learned his psalms from a glossed psalter would have imbibed a commentary along with the text. Benedict recommended the Scriptures as "rectissima norma vitae humanae," and that recommendation was bracketed with references to the "holy Fathers."9
6 "Totus psalmus iste moralis est, et velut cura quaedam et medicina humanae animae datus, cum peccata nostra arguit, et edocet nos secundum legem vivere": Homilia prima, PG 12:131921. The Greek text has been lost; only the Latin version of Rufinus survives. 7 "The Exposition and Exegesis of Scripture from Gregory the Great to St. Bernard," Cambridge History of the Bible, 2:196. 8 Beryl Smalley, The Study of the Bible in the Middle Ages, 2nd ed. (Notre Dame, 1964), pp. 3782. References to the patristic sources can be found in the works of Rondeau and Solignac mentioned in n. 4 above.. 9Regula Benedicti 73; RB 1980: The Rule of St. Benedict in Latin and English with Notes, ed. Timothy Fry et al. (Collegeville, 1981), pp. 294-96. In this chapter Benedict appears to quote from Ambrose's Expositio de psalmis.

The Singing of Psalms It would be difficult to overestimate the power of the psalms in the lives of those who prayed and sang the Office. As the words of the psalms were sung to the prescribed tones, each monk supplied them with his own "jeu spontane des associations, des rapprochements et des comparaisons," a private exegesis of the sacred text.10 The psalms were given a specifically Christian application by the tituli psalmorumand psalm collects. These titles were not those of the Hebrew psalter, but suggestions as to how each psalm should be interpreted in a doctrinal, prophetic, or moral way. Six principal series of such titles exist, and in many of them the interpretation of the psalmist's words as the "vox Christi" is emphasized." The psalm collects were actually used in the Office, though perhaps not widely, until the eighth century. The passage from Cassian's Institutes quoted below is an example of how they were joined to the singing of the psalm.12 The preceding considerations only begin to suggest the centrality of the psalter in the monastic world. In cenobitic communities the recitation of the psalms in common was regulated either by written rule or by the directives of the local abbot.'3 Most of the early monastic rules present rather loose guidelines about the number of psalms required and their distribution through the day or week. St. Benedict was the first to prescribe an ordering in specific detail. When his rule superseded older monastic traditions in the ninth century, its plan for the weekly recitation of the psalter became the norm of the monastic Office. The psalms, in contrast to the prayers and readings of the Office, were performed nearly always in a manner resembling singing rather than nonmusical recitation. Though one can imagine a continuum stretching from stylized public reading to genuine melody, the patristic and medieval texts which mention psalms and readings in succession make a clear distinction between the two. Isidore of Seville says that "a lectio is so called because it is not sung like a psalm or hymn but merely read."'4 The Rule of Sts. Paul and
'Jacques Dubois, "Comment les moines du moyen age chantaient et goutaient les Saintes Ecritures," Le moyenage et la Bible, ed. Pierre Riche and Guy Lobrichon (Paris, 1984), p. 262. 1 Pierre Salmon, Les "Tituli Psalmorum"des manuscritslatins, Etudes liturgiques 4 (Paris, 1959); Salmon discusses each series and gives representative examples in L'Officedivin, pp. 115-23. On the interpretation of the psalms as "vox Christi" see Rondeau, Les commentaires, vol. 2 passim. 12 Andre Wilmart and Louis Brou, The Psalter Collectsfrom V-VIth Century Sources, Henry Bradshaw Society 83 (London, 1949); an improved text has been published by Patrick Verbraken, Oraisons sur les cent cinquante psaumes: Texte latin et traductionfrancaise de trois series de collectes psalmiques, Lex orandi 42 (Paris, 1967). For an interpretation see F. Vandenbroucke, "Sur la lecture chretienne du psautier au Ve siecle," Sacris erudiri 5 (1953), 5-26. 13 For a historical overview of the monastic Office in the West see Paul Bradshaw, Daily Prayer in the Early Church, Alcuin Club Collections 63 (London, 1981), pp. 124-29; and Robert Taft, The Liturgy of the Hours in East and West (Collegeville, 1986), pp. 93-140. Structural questions are paramount in the classic study of Odilo Heiming, "Zum monastischen Offizium von Kassianus bis Kolumbanus," Archivfiir Liturgiewissenschaft (1961), 89-156; and in Corbinian Gindele, "Die 7 Struktur der Nokturnen in den lateinischen Monchsregeln vor und um St. Benedikt," Revue benedictine64 (1954), 9-27. 14 "Lectio dicitur quia non cantatur, ut psalmus et hymnus, sed legitur tantum": Liber etymosive Originum libri XX, 2 logiarum 6.19.9, ed. W. Lindsay, Isidori Hispalensis episcopiEtymologiarum


The Singing of Psalms 539 Stephen, probably from the sixth century, quotes a famous injunction from the Rule (Praeceptum)of Augustine that texts intended to be spoken should not be embellished "with musical figures and the art of melody." The common forms of psalmody, responsorial and antiphonal, consisted of two components: the text of the psalm itself and the refrain inserted between verses. Since there can be no doubt that this refrain was sung, it is only reasonable to suppose that the psalm was presented in a style consistent with the melody of the refrain. The term "recitative," though anachronistic, suggests the manner of performance. Parallels between the formulae used to recite the psalms and epic poetry, folksong, and the more elaborate pieces in the chant repertoire have been proposed by several scholars.15 The first substantial evidence of a theory designed to organize the singing of psalms appears in the Musica disciplina of Aurelian of Reome, a treatise thought to have been written about the middle of the ninth century.'6 Of somewhat earlier date (ca. 835) is the hypothetical archetype of a comprehensive tonary from Metz, which lists the antiphons of the Office and prescribes specific psalm-tone formulae appropriate to each of them.17 Both of these documents concern a body of music which came to be known as "Gregorian" chant. The structural similarities between the psalm-tone formulae of Gregorian chant and those of the local Roman repertoire known as "Old Roman" allow us to trace the history of these formulae back to the eighth century, and thus to a period before the two chant traditions separated. Both Gregorian and Old Roman manuscript sources indicate the psalm tone to be used with a particular antiphon in the same way: the last six notes of the formula used to recite the psalm verses are set above the letters euouae (saeculorumamen).'8 While the earliest treatises employ several terms (varietas, divisio, diffinitio, and differentia) to describe these cadences, eventually the term differentia became the accepted designation for the cadential gesture which linked the psalm verses with a recurrent antiphon. Ambrosian chant employs a similar system. There is a large corpus of such "differences." The need for so many of them has never been satisfactorily explained, nor is it
vols. (Oxford, 1911); PL 82:252. Cf. Tertullian: "the Scriptures are read [and] the psalms are sung": De anima 9, ed. A. Reifferscheid and G. Wissowa, CSEL 20 (Vienna, 1890), p. 310. 15 See Ewald 22 Jammers, "Der Choral als Rezitativ," Archivfiir Musikwissenschaft (1965), 14368; Janka Sendrei, "Beitrage zu den musikgeschichtlichen Beziehungen des volksmusikalischen Rezitativs," Studia musicologicaAcademiaescientiarumHungaricae 13 (1971), 275-88. An illuminating study of the liturgical lectio is Gino Stefani, "La recitatione delle letture nella liturgia romana antica," Ephemeridesliturgicae 81 (1967), 113-30. 16 Musica disciplina, ed. Laurence Gushee, Corpus scriptorum de musica 21 (Rome, 1975). 17 Walter Lipphardt, Der karolingischeTonar von Metz, Liturgiewissenschaftliche Quellen und Forschungen 43 (Mtinster, 1965). 18 Aurelian, Musica disciplina, passim. Regino of Prtim, in his description of the method he followed in preparing a tonary based on the antiphoner of Trier, equated the differentiaewith the "divisiones tonorum" under which he grouped the antiphons (see n. 47 below). To date the Amen and only extensive discussion of the differentia phenomenon is Clyde Brockett, "Saeculorum Differentiae: Practical versus Theoretical Tradition," Musica disciplina 30 (1976), 13-36. On some philosophical implications of the term see Eleonore Stump, Boethius'sDe topicisdifferentiis(Ithaca, 1978), pp. 248-61.

The Singing of Psalms always clear why one pattern makes a more appropriate link with the antiphon than another. The middle of the psalm verse is marked by a mediant cadence (not in Ambrosian chant, however), and the end of the antiphon is linked to the reciting pitch of the psalm tone by a short initium. The typical structure of a psalm tone may be examined in Examples 8-.13 below. The number of times the antiphon was intercalated into the body of the psalm text varied according to circumstances which may never be fully understood.19 There can be little doubt that the antiphon was sometimes repeated by the choir after every psalm verse, a practice which prolonged the Office considerably. This custom is perhaps implied by John Cassian's remark (ca. 415) about "these same [psalms] prolonged by the melodies of antiphons" ("hos ipsos antiphonarum protelatos melodiis").20 A similar inference may be drawn from statements in two important monastic rules of the early sixth century, the rules of the anonymous Master and of St. Benedict. Both permit the omission of antiphons in order to alleviate the burden of the Office on very small monastic communities.21 This dispensation would have been meaningless were not frequent repetition of the antiphon regarded as the norm. According to a tenth-century vita of St. Odo (880-942), second abbot of Cluny, the insertion of antiphons after every psalm verse helped to fill up the long vigils observed by Gallic monks during the long nights of winter.22 The same frequency of repetition is implied by Amalarius of Metz in his
19The principal treatments of antiphonal psalmody from a historical perspective are Henri Leclercq, "Antiphone (liturgie)," Dictionnaire d'archeologiechretienneet de liturgie 1/2:2282-2319; Louis Petit, "Antiphone dans la liturgie grecque," ibid., cols. 2461-88; Jacques Hourlier, "Notes sur l'antiphonie," in Wulf Arlt et al., eds., Gattungen der Musik in Einzeldarstellungen:Gedenkschrift Leo Schrade (Bern, 1963), pp. 144-91; Bruno Stablein, "Antiphon," Die Musik in Geschichteund Gegenwart,1:523-45; Michel Huglo, "Antiphon," The New GroveDictionaryof Music and Musicians, 20 vols. (London, 1980), 1:471-81. For a general background see Thomas Connolly, "Psalm: Latin Monophonic Psalmody," ibid., 15:322-32. Hereafter The New Grove will be cited as TNG. 20Institutes 2.2, ed. Michael Petschenig, Johannis Cassiani De institutis coenobiorumet de octo principalium vitiorum remediis,CSEL 17 (Vienna, 1886), p. 18; English translation by Edgar C. S. Gibson, The Institutes ofJohn Cassian, Nicene and Post-Nicene Fathers, ser. 2, vol. 11 (New York, 1894), p. 205. 21 See the Regula Magistri 55.7: "psalmos vero directaneos dicens, ut supra diximus, urgentem laboris operam" ("moreover, he may say the psalms straight through because of urgent work, as we have said above"). La regle du Maitre, ed. Adalbert de Vogiu, Sources chretiennes 106 (Paris, 1964), p. 260; the English translation is by Luke Eberle, The Rule of the Master, Cistercian Publications 6 (Kalamazoo, 1977), p. 220. Hereafter Sources chretiennes will be cited as SC. The Regula Benedicti (17.6) also allows an exception at terce, sext, and none: "Si maior congregatio fuerit, cum antiphonas, si vero minor, in directum psallantur" ("If the community is rather large, refrains are used with the psalms; if it is smaller, the psalms are said without refrain"): RB 1980, pp. 212-13. 22 "Verum quia eiusdem officii antiphonae, uti omnibus patet, breves sunt, et eius temporis longiores noctes; volentes officium ad lucem usque protendere, unamquamque antiphonam per singulos psalmorum versus repetendo canebant": Vita S. Odonis 10, PL 133:48. See the discussion of monastic vigils in Adalbert de Vogiu, La regle de saint Benoit, 6 vols., SC 181-86 (Paris, 197172), 5:452-63; on, time in the monastic world see J. Biarne, "Le temps du moine d'apres les premieres regles monastiques d'Occident (IV-VI siecles)," in Le tempschretiende lafin de l'antiquite au moyen age (III-XIIIe siecle), Colloque international du Centre national de la recherche scientifique, Paris, 9-12 mars 1981 (Paris, 1984), pp. 99-128.


The Singing of Psalms 541 description of nocturns, part of which consisted of "six antiphons which the choirs repeat alternately through [i.e., between] each verse."23 The last remnants of this older custom did not disappear until the later Middle Ages. "Triumphing" the antiphon for the Benedictusand the Magnificat was one of the last reminders of this solemn psalmody of an earlier time. The etymology of the term (tresltrium-fari,to say three times) describes the practice well. The antiphon was sung thrice: (1) before the Gloriapatri, (2) before the Sicut erat, and (3) after the Sicut erat. Rubrics ("hodie antiphonamus") in the Old Roman Antiphoner of St. Peter's may point to a similar usage in singing the ordinary psalms of the Office.24 In Gregorian chant the singing of the antiphon was later restricted to the beginning of the psalm and to the end of the doxology which closed each psalm. Even this was eventually curtailed, however, when the initial statement became reduced to a single phrase of the antiphon.25 Our present knowledge of Gregorian psalmody derives almost exclusively from theoretical sources: medieval writers on music, the (often anonymous) tonaries, and modern chant books like the Liber usualis. While many of the treatises and a few of the tonaries have been edited and studied by musicologists, the practical tradition preserved in the manuscript antiphoners used in the singing of the Office has received almost no attention.26 Music theorists, whether modern or medieval, seek to discover in a given repertoire a regularity susceptible to description in the form of universals. They not infrequently create and eventually succeed in imposing a uniformity which may
"... ex senis antiphonis quas vicissim chori per singulos versus repetunt": Liber de Ordine Antiphonarii3.4, ed. Jean Michel Hanssens, Amalarii episcopiOpera liturgica omnia, 3, Studi e Testi 140 (Vatican City, 1950), p. 24. That the antiphon could be repeated within the psalm seems to be understood by the tenth-century Commemoratio brevis de tonis et psalmis modulandis when it recommends that the "repetitio antiphonarum quae in fine versuum inter captandum fit eadem qua psalmus celeritate percurrat" ("the repetition of the antiphons which occur between the verses should be at the same speed as the psalm"): ed. and trans. Terence Bailey, Ottawa Mediaeval Texts and Studies 4 (Ottawa, 1979), pp. 106-7. Aurelian of R6eme memorializes a custom of singing the "alleluia" refrain of the laudes (Psalms 148-50) of the Sunday morning Office between each psalm verse: Musica disciplina 20.39-40, ed. Gushee, p. 133. 24 See Charles Du Cange, Glossariummediae et infimae Latinitatis, ed. Leopold Favre, 10 vols. (Niort, 1883-87), s.v. "triumphare," 8:190. A long rubric for the Old Roman psalmody, found on fol. 25 of the St. Peter's antiphoner, will be analyzed by Edward Nowacki, "The Performance of Office Antiphons in Twelfth-Century Rome," Studia musicologicaAcademiaescientiarumHungaricae, forthcoming. Clyde Brockett reports on an example from Mozarabic chant of an antiphon repeated within the psalm: Antiphons, Responsoriesand Other Chants of the Mozarabic Rite, Musicological Studies 15 (New York, 1968), p. 133, ex. 11. 25 Two noted psalters included in the present study (Bibl. Vat., Chigi C.VI.163, and Bibl. Vat., Archivio di San Pietro E 14) prescribe a truncated first statement of the antiphon. 26 Hugo Berger, Untersuchungenzu den Psalmdifferenzen,Kolner Beitrage zur Musikforschung 37 (Regensburg, 1966), collates the psalm tones from selected antiphoners representative of regional German practices. Tables of differentiaehave been extracted from individual antiphoners by modern authors: Willibrord Alfons Heckenbach, Das Antiphonar von Ahrweiler, Beitrage zur rheinischen Musikgeschichte 94 (Cologne, 1971); Paleographiemusicale 9 (Solesmes, 1905-9), p. 2 (= Lucca, Biblioteca capitolare 601), and 12 (Solesmes, 1922-24), pp. 126-54 (= Worcester, Cathedral Library, F.160). The Antiphonale monasticum(Tournai, 1934), pp. 1210-19, and the Liberusualis (many editions) reproduce the (30) most common differentiae.Hereafter Paleographie musicalewill be cited as PM.

The Singing of Psalms not have existed in practice. The medieval theorists who grappled with the challenge of defining how antiphons should be connected to psalm tones encountered difficulties in establishing universal principles. Their work went on for many centuries before standardization was achieved. One of the products of this theoretical activity was a species of handbook known as a tonary.27 The large tonaries divide up the entire corpus of antiphons according to the psalm-tone formulae suitable to each. One of the goals of the classification is to create a smooth musical transition between the psalm tone and its antiphon-refrain. Tonaries summarize in practical form decisions made on the basis of theory. Though some of them include or are attached to brief commentaries, they are usually silent or imprecise about the principles applied to the generation of the catalogue. The compilers of tonaries intended to exercise an influence on practice, as indeed they did if one can judge from the antiphoners, noted breviaries, and noted psalters included in the present study. Though these practical manuscripts postdate by many years the intervention of a normative music theory, they preserve important information about the singing of psalms in the medieval Office. They also permit certain inferences to be drawn about psalmodic customs antecedent to the ones they record. 542 Before discussing the manuscript transmission of differentiaeand other matters related to medieval psalmody, however, I would like to review the ways in which the psalms were rendered in the early monastic Office. Since the codification of the monastic Office began centuries before the appearance of the first antiphoners, literary evidence must be the primary source of information.28 The earliest statements about the singing of psalms in the western monastic Office are found in the Institutesof John Cassian (ca. 360-ca. 435), an eastern monk who had passed his youth and maturity in the monastic settlements of Egypt and the Near East. When he came to Gaul about 415, he found already established there a monastic Office of psalms familiar to him from the East. In the Institutes Cassian does not point to any significant difference between the style of psalm singing practiced in fifth-century Gaul and the customs he had witnessed in Egypt, Palestine, or Mesopotamia.29 This is clear even in an observation which notes a slight variation in the use of the doxology. That practicewhich we have observed in this province [Gaul]- that one sings the psalm, at the conclusion of which all rise and sing with a loud voice: Gloria patri et filio et spiritui sancto - we have never heard anywhere throughout the East, but there, while all keep silence when the psalm is finished, the prayer that follows
The classic study is Michel Huglo, Les tonaires (Paris, 1971). English translations of many representative texts (with the exception of monastic regulations) may be found in McKinnon, Music in Early ChristianLiterature. Psalmody in the monastic rules and the transition from solo to choral singing of the psalms are discussed in my "Monastic Psalmody of the Middle Ages," Revue benedictine99 (1989), 41-74. Some of the essential points are summarized in the current discussion. 29 Institutes, especially books 2 and 3, CSEL 17:16-45.
28 27

The Singing of Psalms 543 is offered up by the singer. They add this hymn in honor of the Trinity only to the end of antiphons.30 In Gaul the doxology was sung by all the monks at the conclusion of every psalm, while the eastern monks restricted its singing to the end of antiphons. The rendition of the psalm itself does not differ: one monk sang the psalm and added the psalm prayer while all listened in silence. Apparently it was not customary in Gaul to attach this psalter collect to the psalm. The monk Rufinus, a contemporary of Cassian, confirms that this custom of solo chanting of the psalm was practiced among the semianchoritic monks of Lower Egypt. In a passage he added to his translation of the Greek Historia monachorumin Aegypto he claimed that "it is the custom there for all to sit while the psalm is recited by one; the others either listen or respond."31 About a hundred years later, the monk and bishop Aurelian of Aries (d. 551) prescribed that the psalms at nocturns were to be divided up among four to six "soloists," each of whom was charged with the singing of two ordinary psalms followed by one with an alleluia refrain.32 In central Italy the Rule of the Master mandated that "a brother who was rebuked [for tardiness] in the
oratory . . may on no account sing a psalm or a responsory or a lesson until

he has made satisfaction."33 The logical assumption in all of these cases (and elsewhere in the Rule of the Master) is that the singing of a lesson, an ordinary psalm, or a responsory was an individual activity, and that a delinquent monk had to forgo his privilege of standing before the community to sing the psalm alone. Many other monastic rules from the fifth and sixth centuries attest to the persistence of solo psalmody in the monastic Office.34 Nothing in the widely observed Rule of St. Benedict (ca. 530) specifically contradicts this traditional monastic practice. Even the plural form found in a famous phrase from the rule, "let us stand to sing the psalms in such a way that our minds are in

30 "Illud etiam quod in hac provincia vidimus, ut uno cantante in clausula psalmi omnes adstantes concinant cum clamore, gloria patri et filio et spiritui sancto, nusquam per omnem Orientem audivimus, sed cum omnium silentio ab eo, qui cantat, finito psalmo orationem succedere, hac vero glorificatione trinitatis tantummodo solere antiphona terminari": Institutes 2.8, CSEL 17:24, translation by Gibson, The Institutes, p. 208. Cassian's use of antiphona as a neuter plural is exceptional; see Bernard Botte, "Antiphona," Sacris erudiri 4 (1952), 239-44. 31 "Moris est autem inibi sedentibus cunctis ab uno dici psalmum ceteris vel audientibus vel respondentibus": Historia monachorum29, PL 21:454. The seated posture of the monks may imply responsorial psalmody in this instance. 32 "Quatuor fratres binos psalmos et alleluiaticum tertium dicant": "Regula S. Aureliani Arelat.," ed. Lukas Holste and M. Brockie, Codex regularum monasticarumet canonicarum(Augsburg, 1759), 1:152. There is a similar instruction for the office of duodecimaon Easter evening (Codex regularur, loc. cit.). 33"Nam frater qui correptus in oratorio fuerit, ... tamen psalmum et responsorium vel versum non imponat": Regula Magistri 73.17, SC 106:310, trans. Eberle, p. 238. 34Some of the evidence for solo psalm singing is less direct and makes sense only in the context defined above. Caesarius of Aries (470-542), Aurelian's predecessor as bishop, instructed women religious that they should not engage in any work or speak during the psalmody ("dum psallitur, fabulari omnino vel operari non liceat"): Regula virginum 10, ed. Germain Morin, S. Caesarii Arelatensis opera varia, 2 vols. (Maredsous, 1942), 2:104. This injunction could be interpreted to mean that their lips were not otherwise occupied.

The Singing of Psalms harmony with our voices" ("sic stemus ad psallendum ut mens nostra concordet voci nostrae"), agrees with comparable expressions of a general exhortatory nature in other rules. In fact, it is a paraphrase of an injunction from the Rule of the Master: "we must cry out to God not only with our voices, but with our heart as well" ("non solum vocibus, sed et, corde ad Deum clamare").35This solo psalmody, common to such widely separated areas, was normative in early-medieval monasticism in the West, just as it was in the East. The order in which each monk sang "his" psalm was governed, not by talent or special office, but by seniority within the community. Thus, the solo psalmody of the ancient monastic Office did not require a specially trained cadre of cantors. Every monk knew all the psalms by heart and, unless mitigating circumstances dictated otherwise, every monk was expected to take his turn as soloist. Respect for seniority in the singing of the psalms, as in all aspects of monastic life, can be traced back uninterruptedly to early monasticism. It was emphasized in the material which reflects the customs of the first true cenobitic communities, those founded by Pachomius in the early fourth century. A typical formulation of this rule of seniority may be found in the fifth-century Rule of the Four Fathers: "Let no one among those assisting at prayer presume to utter the praise of a psalm without the command of him who presides. That ordering is to be maintained, so that no one may presume to precede another of higher rank in the monastery for standing or for the order of the singing of the psalms."36 Seniority is likewise honored in the Rule of St. Benedict, which reads: "Therefore, when the monks come for the kiss of peace or for Communion, or when they sing the psalmody or take their place in choir, they do so in the order decided by the abbot or already existing among them."37 It might be objected that since Benedict does not distinguish here between responsorial and antiphonal psalmody, he could be referring to the verses of responsories, which were always rendered by a soloist. I believe that this quotation must be taken in the context of monastic traditions established by the abbots who wrote before Benedict. They supposed that a single monk would sing the text of the main corpus of antiphonal psalmody as his brethren listened in silence or interrupted him occasionally with a refrain. A further 544
35 Regula Benedicti 19.7, RB 1980, pp. 216-17. Cf. the Regula Magistri 47: "Non solum vocibus, sed et corde ad Deum clamare": SC 106:216, trans. Eberle, p. 207. 36 "Astantibus ergo ad orationem, nullus praesumat sine praecepto eius qui praeest psalmi laudem emittere. Ordo iste teneatur ut nullus priorem in monasterio ad standam vel psallendi ordinem praesumat praecedere" (cap. 6). Edition and commentary in Jean Neufville, "Regle des IV peres et seconde regle des peres," Revue benedictine 77 (1967), p. 77; Neufville's text is reproduced with an English translation in Carmela Vircillio Franklin, et al., Early Monastic Rules: The Rules of the Fathers in the Regula Orientalis (Collegeville, 1982), pp. 20-21. Both the vicinity of Rome and the orbit of the monastery of Lerins in southern Gaul have been proposed as points of origin. 37 "Ergo secundum ordines quos constituerit [abbas] vel quos habuerint ipsi fratres sic accedant ad pacem, ad communionem, ad psalmum imponendum, in choro standum" (63.4): RB 1980, pp. 278-79.

The Singing of Psalms 545 argument can be adduced to support this assertion. Unlike the Master, who counted the responsorial psalms as part of the daily psalmody, Benedict subordinated the responsoriumto the reading which preceded it. He did not reckon the responsories following the readings as part of the weekly obligation to recite all 150 psalms. The oldest psalm responsories, sung during the period between Epiphany and Septuagesima, are not independent, for they rely on the antiphonal psalmody of the preceding nocturn for their texts.38 Thus it is unlikely that Benedict would have spoken of "singing the psalmody" in reference to the solo verses of the psalm responsories rather than to the (solo) antiphonal psalmody. Choral psalmody was not unknown in early monasticism outside the context of the Office, but it was restricted to a few well-defined circumstances. As the monks moved about as a group, from the refectory to the church, for example, they chanted Psalm 50.39 Choral psalmody was customary at the death and burial of a member of the monastic community, but this practice was also observed by devout Christians in secular society. The singing of psalms - obviously no more than a selected few - by large congregations of lay men and women is frequently mentioned with approval by the Fathers.40 Before the last half of the eighth century the monastic literature records a single exception to the practice of solo psalmody in choro (i.e., in the Office). It is found in the Rule of Sts. Paul and Stephen, a central Italian rule which has been dated in the mid-sixth century. Let the senior members of each choir of singers begin the psalm verses; ... after this beginning let all presentlyjoin in together - if this is possible - on the first or second syllable and as if from a single mouth, that there may be no disorder among the singers, something which often happens, particularlyas the result of a confused beginning and a certain self-willeddissension.41 This isolated evidence, if indeed it does refer to choral rendition of psalm verses in the Office, would be the only monastic document before the late
38 See Heiming, "Zum monastischen Offizium," pp. 134-35; Petrus Nowak, "Die Strukturele26 mente des Stundengebets der Regula Benedicti," Archivfur Liturgiewissenschaft (1984), 27486; Raymond Le Roux, "Les repons 'de psalmis' pour les matines de l'Epiphanie a Septuagesime selon le cursus romain et monastique," Etudes gregoriennes6 (1963), 39-148. 39 According to decrees prepared in 816 for a council on monastic life held at Aachen, this psalm was to be sung "choris alternantibus": "Actuum praeliminarium Synodi I. Aquisgranensis commentationes sive Statuta Murbacensia" 3, ed. Kassius Hallinger, Corpus consuetudinummonasticarum, 1 (Siegburg, 1963), p. 443. 40McKinnon, Music in Early ChristianLiterature,passim. 41 "Initium versuum psallentium in choro priores qui in eis stant incipiant; ... quibus incipientibus mox omnes, si potest fieri, in prima aut secunda syllaba pariter unanimiter et uno ore subjungant: ut non sit dissonantia cantantium, quae maxime ab inordinato initio, et quodammodo contentiosa varietate solet accidere" (cap. 5), in J. Evangelista Vilanova, Regula Pauli et Stephani: Editi6 critica,i comentari,Scripta et Documenta 11 (Montserrat, 1959), p. 110. Adalbert de Vogue dates this rule in the second half of the sixth century: Les regles monastiquesanciennes (400-700), Typologie des sources du moyen age occidental 46 (Turnhout, 1985), pp. 13 and 58.


The Singing of Psalms

eighth century which points to choral psalmody in that context.42 The manuscript tradition of the Rule of Sts. Paul and Stephen does not begin until the ninth century, however, and one is justified in wondering whether the guidelines for choral singing of the psalm verses were part of this rule three centuries before. By the ninth century choral psalmody seems to have become more common, and instructions similar to those found in the Rule of Sts. Paul and Stephen occur in various pieces of Carolingian monastic legislation. The document known as Memorialequaliter (II) directs that "when you are singing psalms in choir, do so with a harmonious and concordant voice [consona et concordi voce]; let those who are best able to do so begin the verse, so that the rest can join in on the first or second syllable."43From this relatively late date on, the evidence for choral psalmody as the norm for the celebration of the Office becomes stronger, though deeply entrenched local customin this as in other matters - would not have been readily surrendered. The causes which in the last half of the eighth century stimulated the transition from solo to choral psalmody in the monastic Office cannot be accurately determined. Since memorization of the psalter had always been a monastic obligation, the conditions for choral psalmody had been present for centuries. Its actual introduction, however, represented a fundamental change in monastic spirituality: the monk no longer meditated on the sacred text, but prayed it himself. He exchanged quiet "rumination" on the text for a more active involvement in the opus Dei. This development had musical implications which could not be ignored. Before the advent of choral psalmody considerable latitude could be allowed to the individual monk in the singing of his psalm. As long as he presented the other members of the community with a clear musical signal when the antiphon was to be sung, he could have freely varied the melodic formula to which the psalm was set.44 There must have been agreed boundaries, but one suspects that in practice highly diversified formulae were found from one monastery or diocese to another.

The simultaneous singing of the psalm verses by a large group of untrained singers demanded the development of some general principles that could be followed by the entire choir. Instead of the liberty allowed a solo psalmist in the choice of formula and text adaptation, there was now urgent need for general agreement on the ways all 150 psalms could be sung chorally and linked with a changing repertoire of antiphons. The entire choir would have
42 Corbinian Gindele interprets "versus," somewhat implausibly, as a refrain: "Doppelchor und Psalmvortrag im Fruhmittelalter," Die Musikforschung6 (1953), 298. 43"Quando in choro ad psallendum statis, consona et concordi voce psallite, et illi incipiant versus qui prae ceteris utilius possunt, ut ad primam syllabam vel secundam ceteri convenire possint": Corpus consuetudinummonasticarum,1:253. 44The Commemoratio brevis presumes that when the antiphon is inserted, it will be sung at the same tempo as the psalm itself; ed. Bailey, p. 106. The Commemoratio does not imply that the antiphon was repeated after every verse, i>owever.

547 The Singing of Psalms to select the same cadence formulae from the many possibilities available. Standardization and discipline became high priorities. Since the choice of differentia rested on musical considerations, it did not take into account the variable accent patterns of the end of each verse and half-verse of the psalm. The solution found in the antiphonal psalmody of the Office places the accents of the text invariably on the same pitches of the differentia.This "accentual" method requires the repetition of certain pitches, an adjustment more difficult for a choir than the alternative method, simple assignment of the last six syllables of the verse to the six pitches of the differentia.Many scholars believe that originally there was no such adjustment and that the fixed (or cursive) cadence prevailed.45 It is impossible to establish the order of priority from the manuscripts under discussion here. The small number of comparisons they permit indicates that the differentiae were adapted to the accent patterns of the text. Though the few written-out examples I have discovered (Exx. 11-13 below) are set to nonpsalmic texts, there is no reason to believe that psalm texts were treated differently. The change from solo to choral recitation of the psalms did not take place quickly or without incident. Even a century after the transition began, it proved to be (at least in part) the source of problems at Trier. In a letter (ca. 900) to Bishop Rathbod of that city, Regino of Prum observed that in certain churches of the diocese of Trier the "chorus of psalm singers resounds with discordant voices" ("chorus psallentium psalmorum confusis resonaret vocibus"). Regino attributed the difficulty to a lack of agreement on the tones to be used in singing the antiphonal psalmody. On the basis of the Trier antiphoner he drew up a tonary which grouped the antiphons according to the "divisions of the tones, that is, the differentiae"("divisiones etiam tonorum, id est differentias") proper to each. In a letter prefaced to the tonary he summarized his editorial principles: respect for tradition, organization of the psalm formulae according to harmonica disciplina, and a reduction in the number of differentiae by the omission of those he deemed "superfluous." Lest he be reproached by less experienced musicians ("superstitiosis musicis"), he placed the superfluous differentiaein the margins of the pages.46 This must
45 See Terence Bailey, "Accentual and Cursive Cadences in Gregorian Psalmody," Journal of brevisBailey the AmericanMusicological Society29 (1976), 464-71; on the basis of the Commemoratio points to a mixed practice for the mediant cadences (p. 469). See also Don Randel, "Antiphonal Psalmody in the Mozarabic Rite," InternationalMusicological Society:Report of the Twelfth Congress, Berkeley,1977, ed. Daniel Heartz and Bonnie Wade (Kassel, 1981), pp. 414-22. Byzantine psalm formulae did not adapt to changing accent patterns according to Oliver Strunk, "The Antiphons of the Octoechos," Journal of the American Musicological Society 13 (1960), 50-67, reprinted in Strunk's Essays on Music in the Byzantine World (New York, 1977), pp. 165-90, especially p. 174. 46 "Cum frequenter in ecclesiae vestrae dioecesibus chorus psallentium psalmorum melodiam confusis resonare vocibus, propter dissonantiam toni, et pro huiuscemodi re vestram venerationem saepe commotam vidissem; arripui antiphonarium, et eum a principio usque in finem per ordinem diligenter revolvens, antiphonas, quas in illo adnotatas reperi, propriis, ut reor, distribui tonis; divisiones etiam tonorum, id est differentias, quae in extrema syllaba in versu solent fieri, ut decens et conveniens fiat concinentia, sicut a maioribus nostris traditae sunt, et sicut ipsa harmonicae disciplinae experientia monstravit, distinctis ordinibus inserere curavi. Adiiciunt autem quidam et alias divisiones, quas superfluas arbitramur. Sed ne a superstitiosis musicis


The Singing of Psalms

have led to their demise, for the extant sources of his tonary do not record these extra differentiae. Based on the evidence of the admittedly defective edition of Regino's tonary published by Coussemaker, it appears that Regino was extremely selective in choosing differentiae for inclusion. He fits all the antiphons of the repertoire to about half the differentiae(twenty-eight vs. fiftyfive) found in the slightly earlier Carolingian tonary of Metz.47 Regino's methodology offers a valuable insight into the principles that other, anonymous compilers of tonaries may have followed. Especially interesting is his critical stance toward the tradition and his readiness to discard certain formulae. A century later, the author of the Commemoratio brevis was less inclined to reductionism: he tried to include many traditional nuances, but he was not able to construct an entirely coherent synthesis of material "de diversis collecta." Indeed, the diversity of practice - though in this case not necessarily of differentiae - presented challenges which seem to have overwhelmed the author, for his manual is both incomplete and repetitive. I believe that many of his problems can be traced back to the difficulties he encountered in trying to adapt idiosyncratic solo psalmodic formulae for choral participation.48 Although the eleventh-century tonaries are still comparatively rich in differentiae (see Appendix B), later medieval theorists took a position which emphasized the need for reduction and revisions which (they thought) would impose better order on the system. In the early twelfth century the tonary in Florence, Bibl. Nazionale, Conv. sopp. F.III.565, justified reductions in the number of differentiaeon the basis of "musicae artis." Written comments in the tonary acknowledge that tone IV has nine differentiae, but maintain that "regulariter et naturaliter" it ought to have only four.49 Similar reductions are prescribed for other tones. John of Afflighem classified all differentiae into three categories: (1) those that were fitting and necessary, (2) those that were fitting but unnecessary, added simply for decoration, and (3) those neither fitting nor necessary. He had little patience with the second category and also implied that the proliferation of differentiaewas undesirable because it had been engendered by "corrupt" antiphons.50 Similar editorial interven-

reprehendamur, eas subtus aut supra in margine adnotare studuimus, periti cantoris iudicio relinquentes, utrum eas necessarias, an supervacuas opinari velit." Epistola de harmonicainstitutione, ed. Martin Gerbert, Scriptoresecclesiasticide musica sacra potissimum,3 vols. (St. Blaise, 1784; repr. Hildesheim, 1963), 1:230-31. 47 Huglo, Les tonaires, pp. 71-89. Regino's tonary is reproduced in Edmond de Coussemaker, Scriptorumde musica medii aevi nova series, 4 vols. (Paris, 1864-76; repr. Hildesheim, 1963), 2:373. The Carolingian tonary has been edited by Walter Lipphardt (see n. 17 above). 48 Bruno Stablein maintains that it would have been impossible for a choir - as opposed to a soloist - to negotiate the setting of verses from Psalm 71 illustrated in the Commemoratio brevis: "Gallikanische Liturgie," Die Musik in Geschichte und Gegenwart,4:1321 and ex. 20. This example corresponds to exx. 53-55 in the Bailey edition of the treatise. 49Huglo, Les tonaires, p. 189. 50 De musica cum tonario 22, ed. Joseph Smits van Waesberghe, Corpus scriptorum de musica 1 (Rome, 1950), pp. 153-56; the treatise has been translated by Warren Babb, Hucbald, Guido, and John on Music, Music Theory Translation Series 3 (New Haven, 1978), pp. 159-61. In this

549 The Singing of Psalms tion caused Cistercian and Dominican liturgical books to deviate from traditional practices. The Tonale S. Bernardi proposed that only three differentiae were required to link the reciting tone to antiphons with either a low, midrange, or high initial pitch.51 A few theorists of the later Middle Ages approached the differentiasystem with radical proposals that would have virtually eradicated it. The twelfthcentury Cistercian monk Guy d'Eu thought that a single differentia would suffice for each mode.52 Elias Salomon (fl. 1274), a secular priest, also shared the view that one saeculorum amen per mode was quite sufficient "de artis natura."53 In the fourteenth century Heinrich Eger von Kalkar (1328-1408) wanted to jettison all of the troublesome "caudas diversas" as modern superfluities.54 Comparable expressions can be found in other theorists. This process of radical reduction of differentiae can be observed strikingly in the psalmody of St. Peter's basilica between the twelfth and the fourteenth century. The Old Roman antiphoner of the basilica (Arch. di S. Pietro B 79) contains a very large number of differentiae(fifty-eight), while the fourteenthcentury Gregorian antiphoner (Arch. di S. Pietro B 87) contains an unusually small repertoire of them (twenty-three and the tonusperegrinus). This cannot be explained entirely by the change in chant repertoire which took place at the basilica in the intervening years.55 The reduction in the number of differentiaein the manuscript tradition was promoted by theorists and by the compilers of tonaries. The theorists never suggested expanding the corpus, just contracting it. This program - perhaps along with the need to simplify the musical tasks of the monastic or collegiate choir - inevitably influenced the contents of the antiphoners. While the old soloistic tradition, exclusively oral, would have fostered maximum diversity, some of this diversity must be attributed to regional variation. In the thirteenth century Petrus de Cruce mentioned that various cities still clung to local customs of psalm singing.56 This observation seems to be
passage John also expresses his preference for differentiaewhich end with a single note rather than with a neume. 51 In Gerbert, Scriptores, 2:269. Cf. the Regulae de arte musica of Guy d'Eu (Coussemaker, Scriptorum, 1:181), on which the Tonale is based. One of the manuscripts which contain the Regulae (Paris, Bibliotheque Ste. Genevieve 2284; 13 c.) claims that many of the differentiaewere created "inconveniently" and work just as well in one mode as in another; Claire Maitre, "Recherches sur les Regule de Arte Musica de Gui d'Eu," Les sources en musicologie, Actes des journees d'etudes de la Societe francaise de musicologie a l'Institut de recherche et d'histoire des textes d'Orleans, 9-11 septembre 1979 (Paris, 1981), pp. 79-86. 52 "Arbitror autem, immo plane affirmo unicuique modo tantum, sed propriam differentiam posse sufficere": Regulae, Coussemaker, Scriptorum,2:182. He nevertheless transmits the traditional teaching. See Norman Smith, "Guy de Cherlieu," TNG, 7:858-59. 53Scientia artis musicae 11, in Gerbert, Scriptores,3:30. 54 ed. Heinrich Huschen (Cologne, 1962), p. 61, as quoted by Berger, UntersuCantuagium, chungen zu den Psalmdifferenzen,p. 151. 55The fact that B 87 begins as a noted psalter may have some relevance: invariably this category of manuscript has fewer differentiae.See the list of psalters in Appendix A. 56 "De differentiis seu principiis eorum, quot differentias seu principia unusquisque eorum habeat, nulla musicae regula numerum certum declaravit. Usus enim civitatum, qui diversi sunt,

The Singing of Psalms confirmed by the information provided by the Italian antiphoners covered in the present study. In order to test the hypothesis that large numbers of differentiae represent the survival of a tradition of solo psalmody, I undertook an investigation of both the quantity and the diversity of differentiaepresent in a representative sampling of medieval Italian antiphoners, both Old Roman and Gregorian.57 Recourse to the Gregorian practical sources was necessary to permit a genuine comparison between Gregorian and Old Roman systems of antiphonal psalmody. Gregorian theoretical constructs - tonaries and treatises - have no Old Roman parallels. Only after the psalm tones had been extracted directly from the Gregorian antiphoners would it be possible to draw reliable conclusions about the relationship between the two traditions with respect to the antiphonal psalmody of the Office. Furthermore, the direct examination of the Gregorian practical tradition promised new insights about the Gregorian system itself. A preliminary survey of antiphoners from various parts of Europe led me to restrict the Gregorian field to Italian antiphoners whose pitches could be accurately transcribed. This offered both a manageable corpus of material and an indigenous repertoire which could be compared with the contents of the Old Roman antiphoners.58 The examination of these manuscripts also provided an opportunity for collecting the rare cases of psalm tones set to a complete text, the better to understand how psalm texts and those of other provenance were fitted to the differentiae. Though complete antiphoners were the most desirable sources, it was sometimes necessary to have recourse to incomplete ones or to a single surviving volume of an original winter-summer pair. This was not always an obstacle, however, since the entire repertory of differentiaecould conceivably be preserved even under these circumstances. Since not all differentiaewere sung throughout the liturgical year, however, those would be lost which were concentrated around a particular feast or season contained in the missing section(s).59 In addition to the antiphoners I examined several noted psalters and noted breviaries. Many of these sources have only sporadic notation of differentiae: the staves drawn to receive them remain empty. This is true of
dant eis differentias diversimodo, tum quia unus plus, alter vero minus": Tractatusde tonis, ed. Denis Harbinson, Corpus scriptorum de musica 29 (Rome, 1976), p. vii. 57 The only surviving Old Roman antiphoners are Bibl. Vat., Arch. di San Pietro B 79 (St. Peter's, 12 c.), and London, Brit. Lib., Add. 29988 (possibly from the Lateran, 12 c.). The first of these is scheduled for publication in Monumenta monodicamedii aevi, edited by Eugene Leahy. 58Helmut Hucke presented the psalm tones of the Old Roman antiphoners in "Karolingische Renaissance und gregorianischer Gesang," Die Musikforschung28 (1975), 4-18. There are also observations on Old Roman psalmody (not always easy to follow, as Hucke points out) in Ewald Jammers, Musik in Byzanz, im pdpstlichenRom und im Frankenreich(Heidelberg, 1962), pp. 12631. 59For example, in the antiphoner Benevento, Bibl. cap. V.22, the equivalent of Gregorian tone I.G does not occur early in the manuscript, but it is used later to the exclusion of all other mode I differentiae. If the later portion of that manuscript had failed to survive, this differentia would have seemed to be absent.


The Singing of Psalms 551 even large breviaries like Monte Cassino 420 and Rome, Bibl. Casanatense 1574.60 Since my principal interest was in quantity, I did not attempt to draw up separate tonaries for the manuscripts in this survey.61 Nor did I compare the modal assignment of antiphons in the antiphoners with the same antiphons in the tonaries. Because of the sometimes subtle distinction between one differentia and another, I decided to rely on manuscripts whose pitches could be read. This meant that manuscripts earlier than the twelfth century could not be included.62 Neumed manuscripts could not provide much control when several differentiaein a single mode might be written in virtually the same way. Even with staff notation the pitches are not always obvious, for the familiar differentiae are on occasion entered so cursorily that the exact pitches intended can be difficult to divine. (Chigi C.V.137, a breviary from Farfa, and Vat. lat. 14446, a breviary from Caiazzo near Naples, are two examples.) Because of this problem differentiaewhich appear to be anomalous are difficult to judge: do they represent archaic tradition or a slip of the pen? I have invariably considered such dubious entries as representing differentiae only if clearly attested elsewhere in the manuscript. This conservative approach avoided the creation of a differentia when no difference was intended. Because of these problems I cannot claim that the statistics presented here on the number of differentiaein a given manuscript are absolutely definitive. Considering the large amount of documentation examined in the course of this project, sheer inadvertence could have easily allowed some formulae to slip by unobserved. The antiphoners, noted breviaries, and noted psalters chosen for this study (about 50; listed in Appendix A) range in date from the twelfth to the fourteenth century and cover most of the Italian peninsula, from Naples and the Beneventan region in the south to Lombardy, Piedmont, and Friuli in the north. The three-hundred-year time frame assured that a sufficiently large number of complete antiphoners would be included; it also allowed for the possibility that older practices might be conserved in relatively late manuscripts.63Both secular and monastic manuscripts are represented; there seems to be no distinction between them with respect to the psalm tones. The Beneventan tradition of psalmody appears to be in some respects a special enclave. Old Roman psalmody most definitely is, while antiphoners from the central and northern areas tend to present the usual Gregorian psalm tones. One regional variant of Gregorian chant exhibits certain features which have
60 Even breviaries without music could have been employed in the sung Office. A rubric in Rome, Bibl. Naz. Centr., Farfa 22, directs the cantor not to begin the psalm with its first words, since these had already been sung as the antiphon (fol. 53v). 61 The lists of differentiaeaccompanying the published facsimiles of Lucca, Bibl. cap. 601 (PM 9), and Worcester, Cathedral Library F.160 (PM 12), were compiled by the editors of these volumes. 62 Only one earlier manuscript, an antiphoner in the Biblioteca comunale at Todi (MS 170, possibly from the end of the eleventh century), was complete enough to be included. 63Jean Claire has discovered such archaic practices in late manuscripts from Aachen (Bibl. cap. 35) and Metz (MS 461): "Les repertoires liturgiques latins avant l'octoechos I: L'office ferial romano-franc," Etudes gregoriennes 15 (1975), 15-16 and passim.

The Singing of Psalms ramifications for the differentiae. In this so-called "Germanic dialect" a pitch with a minor second above it will slide up to the higher tone. For example, the major second a-b becomes a-c, and d-e becomes d-f. This shift occurs in manuscripts from German-speaking lands, from eastern Europe, and in manuscripts from Friuli and the Veneto. It could be asserted - against the view that I am proposing - that a large repertoire of differentiaerepresents merely scribal sloppiness or later diffuseness rather than the heritage of antiquity.64 I cannot offer any definitive proofs because of the relatively late date of the sources, but I can point to an apparent trend toward reduction in the number of psalm-tone formulae. The large number of differentiae represented, albeit sporadically, in twelfthcentury antiphoners is no longer found in antiphoners of the thirteenth and fourteenth centuries. As the differentiaebecome fewer, they tend to be confined to those which make up the core of the common tradition, and local variants tend to disappear. This development is consistent with the information culled from medieval theorists and tonaries. My experience with the manuscripts, moreover, leads me to conclude that it is usually possible to separate intent from carelessness. Some of the manuscripts listed in Appendix A contain liturgical items of varying types: for example, the noted psalter in Archivio di S. Pietro E 14 also contains canticles, hymns, the Office of the Virgin, and the Office of the Dead.65 Several of the antiphoners listed are well known for the tonaries associated with them: Piacenza 65, Vercelli 70, Monza 16/82, and Lucca 603. The antiphoners naturally represent the largest group of sources consulted for the present study, and they yielded the most information on the practice of psalmody. A few of the noted breviaries (Todi 170, Vallicelliana C.13, Benevento V.22) had the psalmody notated consistently, but most of the others contain only sporadic entries of differentiae,even when the staves had been drawn to receive them. Such lack of thoroughness also reduced the value of the noted psalters considerably. The oldest complete Italian source in staff notation for the antiphonal psalmody of the Office seems to be Todi 170, an eleventh-century noted breviary with 40 differentiae- a moderately high figure which corresponds to that found in the antiphoners of this period. A manuscript in the chapter library at Ivrea (62, olim 64) has 45. Two twelfth-century antiphoners from Lucca (Bibl. cap. 601 and the incomplete 603) have approximately the same number (ca. 40). The largest Gregorian source I have been able to discover 552
64See Lipphardt, Der karolingischeTonar, pp. 222-45; and Graduel romain: Edition critique4/1 (Solesmes, 1960), p. 291 ("c'est la partie la plus ancienne de la tradition qui est la plus differenciee, tandis que la partie la plus recente tend vers l'unification"). One can agree with Paul Cutter's observation that "it was just those melodies that were sung almost every day with which the greatest liberties were taken": "The Old Roman Chant Tradition: Oral or Written?" Journal of the AmericanMusicological Society 20 (1967), 173. 65For a list of the contents of this and other Vatican manuscripts see Pierre Salmon, Les manuscrits liturgiques de la BibliothequeVaticane, 1 (Vatican City, 1968); another similarly mixed source is Vat., Chigi C.VI.177 (Salmon no. 149).

The Singing of Psalms 553 comes from the twelfth century: an antiphoner (MS 84) preserved in the archiepiscopal library at Udine in Friuli. This manuscript, from the diocese of Treviso near Venice, contains about 67 differentiae.Unfortunately, almost all the medieval manuscripts which remained at Treviso were destroyed during the Second World War. Not even a catalogue of prewar holdings exists, so it is impossible to determine whether or not the Udine manuscript represents an important local tradition. The Romano-Beneventan tradition, as it existed in the twelfth century, is preserved in four manuscripts from Benevento, three incomplete ones at Monte Cassino, and another incomplete manuscript now at Naples. This tradition seems to have been richly supplied with differentiae (45-55), some of which are not found outside this enclave. This proliferation of differentiae could be related to the absence of a music theory which directly addressed the Beneventan repertoire as well as to the oft-demonstrated inclination of Beneventan manuscripts to retain special practices against the pressure of Gregorian conformity.66 There are, to be sure, smaller collections from the and a group of twelfth century: Vallicelliana C.5 and C.13 (31 differentiae)67 three related manuscripts from Klosterneuburg, witnesses of the "Germanic" chant dialect. Several of the other manuscripts consulted from this period were incomplete and hence difficult to evaluate. Evidence from the late twelfth and early thirteenth centuries is mixed. The only manuscript supplied with an unusual number of differentiae (Vat. lat. 14676) comes from Pavia; it has 50 formulae. The other contemporary antiphoners are unexceptional or are too fragmentary to provide reliable data. All but two (Vat. lat. 14446 from Caiazzo near Naples and Vat., Borg. lat. 405, from central Italy) come from Tuscany (Lucca, Florence) and further north. A fragmentary source from Udine (Bibl. arch. 72) contains a very small repertoire of differentiae. No fourteenth-century manuscript among those collated has an unusual number of them. Two antiphoners which represent the tradition of Aquileia at that late date (Gorizia A and B) include a small repertoire which is in agreement with the common Gregorian tradition. The presence of large numbers of differentiaein certain manuscripts (and in a few tonaries) has not received a satisfactory explanation. In a given mode the differentiae usually share six basic structural tones, corresponding to the six syllables of the closing words of the doxology, "saeculorum amen." Two pairs of tones share the same reciting pitch: a in modes I and IV, c in modes III and VIII. There is never any confusion among them, and I have discovered no instances of a crossover from one mode to the other within these
66For an excellent overview see Thomas Kelly, "Montecassino and the Old Beneventan Chant," Studies in Early Music History 5 (1986), 53-83. Beneventan psalmody will be discussed in Kelly's forthcoming book, The Beneventan Chant. Most earlier scholarly attention, like Dom Hesbert's treatment in PM 14, has been devoted to the chants of the Mass. 67The winter Office only of both manuscripts is catalogued in Jacob Ledwon, "The Winter Office of Sant'Eutezio di Norcia," Ph.D. dissertation, State University of New York at Buffalo, 1986.

The Singing of Psalms pairs. There is still another uniform characteristic: with the exception of tones V and VII, all the tones remain level or descend at the cadence.68 The variety of differentiae arises not only because of their different final pitches, but also because of the embellishment - at least'from the standpoint of the modern analyst - of the basic six-note structure. The purpose served by this embellishment is not immediately clear. Although all the tones with more than a single differentiaor two exhibit this characteristic, those belonging to modes IV and VIII illustrate it particularly well (Ex. 1). 554

Specimens of typical differentiae (a) Mode IV

(b) Mode VIII e e

a a

Q-- .
ea "IJ





* -IUD


* j**V'., 0- -@,,

D e=



There is a modern inclination to perceive similar differentiaeas variants of a prototypical Urform. Such an analysis may seem inviting to twentieth-century ears, because the parameters of variation found among the differentiaewithin a mode or even across modes can be so narrow. While I cannot claim that they should not be reduced to a small number of Urformen (it would be a simple exercise), I can only point out that this kind of reductionism does not represent the perspective of most medieval theorists. They set up separate categories for these differentiaefor use with specific antiphons. Furthermore,


Exceptions are tone I.a in Chigi C.VI.177 and Monte Cassino 420, fol. 66.

The Singing of Psalms 555 some of these "variants" turn up consistently all over Italy and elsewhere in Europe. In most cases the differentia performs its function of linking psalm and antiphon well. This is particularly true when the differentiamakes a felicitous approach to the reinception of the antiphon by foreshadowing or providing a mirror image of its initial notes (Ex. 2).
2 EXAMPLE Linkbetween and differentia antiphon A
Todi 170, fol. 19r e Todi 170, fol. 156r u o u a e Deus au - ri-bus nos-tris...


Al-le-lu - ia, al-le-lu - ia... Monte Cassino 420, fol. 94r Quoniam Naples XVI.A.7, fol. 23r in se - cu-lum...



I. * 11
Es- to mi-chi do- mi-nus...

Monte Cassino 420, fol. 1 lr

' it Vi - vit do - mi-nus... 11

Ivrea 62, fol. 113v



O mors e- ro mors tu a...

Many additional examples could be cited. The need to connect the psalm with its following antiphon cannot account for all the variants in Ex. 1, particularly with respect to the first two pitches of the differentia, rather far removed from the beginning of the antiphon as they are. One can even find in every antiphoner cases in which the connection between differentia and antiphon must be judged somewhat clumsy (Ex. 3). There are many situations, of course, which display a more neutral character and elicit subjective impressions of suitability based on tessitura and intervallic relationships.69 A
69After a thorough'investigation of antiphon assignments in Italian tonaries, Paul Merkley concluded that in the sources he examined "there appears to be no consistent relationship between the saeculorum amen formulas and the incipits of the antiphons assigned to them": "Conflicting Assignments of Antiphons in Italian Tonaries," Ph.D. dissertation, Harvard University, 1985, p. 250.

The Singing of Psalms few formulae have only a minimal cadential inflection from the reciting pitch (Ex. 4; see also Ex. 6). Though this might be determined entirely by musical reasons, one can advance the hypothesis that this stark simplicity might represent an archaic stage of development. The effect resembles that of the tones used for lessons and collects.
Link between differentiaand antiphon Ivrea 62, fol. 4r


i.* * * * .,
e u o u a e

1 I
E runt sig-na...


Brit. Lib., add. 17302, fol. 9r


---_ _^_

Ce -


lum et ter -

Brit. Lib., add. 17302, fol. lOv


in nu - bi-bus...

EXAMPLE4 Differentiae with minimal cadentialinflection (a) Ivrea 62 e (b) Monza 15/79 u o u a

. A. 1
e e

- IIu o u a


P =f = =_ = A 4 --


... !1
-_ 1- - i ? - --X . _ 11 |

(c) Vat. lat. 14676

-C -

Over the centuries the proportion of differentiaeremained remarkably constant across the modes. The psalm tones for modes II, V, and VI have typically only one or two cadential patterns, which are found quite consistently in all regions (Ex. 5). Many manuscripts have but a single differentia for each of these tones. The antiphoner from Treviso (Udine 84) is unusual in having a variety of terminations for the psalm tone of mode II. The lack of variety present in the psalm tones for modes II, V, and VI is as striking as the diversity found among the tones for modes I, IV, and VII, a diversity attested both in the manuscript tradition and in the theoretical



The Singing of Psalms

EXAMPLE5 Differentiae of modes II, V and VI (a) Beneventanmss. (II)



u o ''9 u




_ I

A .I


(b) Casanatense1574 (V)


! -



e u o u a e (c) Ivrea 62 (VI) e u o u * I"


sources. (This is true even apart from the differentiae assigned to the altera positio of tone IV with its final on a.) Could it be possible that a small number of differentiaepoints to a mode which was established as an independent entity later than those modes which have more differentiae? Comparisons between Italian antiphoners and manuscripts from north of the Alps prove that a core repertoire of differentiaewas practically universal in all modes. Certain of them are to be found with remarkable consistency in virtually every manuscript consulted: the tone classified VIII.G in the modern chant books (see Ex. 9: "Ut confirmet") is the best example - only the Beneventan tradition remains aloof (Benevento, V.19-20, 21, 22; Monte Cassino 542; Naples XVI.A.7). There are various regional preferences or adaptations as well, but it is difficult to be dogmatic about the exact content of these regional traditions.70 The total number of differentiae used in the Old Roman tradition (more than 102) far exceeds what can be encountered in the Gregorian tradition. Of all the medieval antiphoners and tonaries consulted, none has a larger number of differentiae than the Old Roman antiphoner now in the British Library. While the exact provenance of this manuscript is somewhat uncertain, it is almost certainly from Rome or its environs, perhaps from the Lateran, as Bruno Stablein believed. It cannot under any circumstances be from St. Peter's, for the contemporary Old Roman antiphoner of the basilica (Arch. di S. Pietro B 79) had a very different custom of psalmody. These two antiphoners disagree thoroughly on the basic repertoire of psalm tones. In other words, the two Old Roman antiphoners differ far more widely in their psalmody than do any two Italian manuscripts selected at random or, one suspects, than do any pair of Gregorian manuscripts regardless of their
70 Zoltan 25 Falvy, "Zur Frage von Differenzen der Psalmodie," Studien zur Musikwissenschaft (1962), 160-73. Several of the differentiae Falvy associates with specific northern traditions turn up in Italy as well.


The Singing of Psalms

geographical origin.71 Thirty-nine differentiaeare found in one or the other manuscript, but not in both. For example, in the equivalent of Gregorian mode VI there are 7 formulae in B 79 (the St. Peter manuscript) and 12 in the British Library antiphoner, yet only 3 of these formiulae are common to both manuscripts. This is an amazing discrepancy, unlike any encountered in the Gregorian tradition - a "regional" practice carried to ultimate lengths by churches on opposite sides of the Tiber. Helmut Hucke has suggested that Old Roman chant borrowed from Gregorian-Frankish chant the system of (eight) church modes and that the corresponding psalm tones were taken over in Rome "spit, unsystematisch, in verschiedenen Redaktionsschuben und unvollstandig, ohne das System theoretisch zu bewaltigen." The system was imposed in its complete form on the Mass psalmody, while in the Office it was incorporated only fragmentarily ("stiickweise").72Hucke points to E-mode antiphons with a psalm tone reciting on the final and to the absence of the equivalent of Gregorian mode V (F final with c recitation) from the Old Roman antiphoners.73 There is no reason to believe, however, that specific Gregorian psalm-tone formulae were imported to Rome along with the octoechos. The more likely explanation is that the Old Roman tradition preserves an original layer of luxuriant psalmodic variety which was subsequently lost to the Gregorian tradition as a result of the activity of theorists and the influence of tonaries.74 This original layer of formulae seems to be common to both traditions, a unique feature which has never been adequately emphasized. The psalm tones are, in fact, the only area in which the two traditions consistently share identical melodies. It does not seem likely that Old Roman chant singers would have felt a need to supplement their already rich repertoire with additional Gregorian formulae.75 The absence of a controlling theory specific to Old Roman chant allowed differentiaeto proliferate: a situation most likely
71 They could never have been used in the same church without creating enormous confusion. This discordance between psalmodic practices strengthens Stablein's hypothesis that the British Library antiphoner could have come from the Lateran (Monumentamonodica medii aevi, 2:30*). The diversity of the two practices could have been one of the reasons why "strenui cantores" were needed at the Lateran in the time of Prior Bernhard (1145) to respond to a city choir at vigils and matins on the feast of John the Baptist. At this period the Lateran canons came, as Bernhard informs us, "ex diversis terrarum partibus." They might have encountered difficulty in adapting the psalms to the rich variety of psalm tones in use at the Lateran. St. Peter's would not have been the best place to recruit these cantors; they must have come from city churches which shared the custom of the Lateran. Bernhardi Cardinalis . . . Ordo officiorumecclesiaeLateranensis, ed. Ludwig Fischer, Historische Forschungen und Quellen 2-3 (Munich, 1916), p. 140. 72 "Karolingische Renaissance," p. 11. 73 The two F-mode antiphons cited by Hucke from the Old Roman antiphoner in the British Library, Ecce iam venit and Haurietis aquas, are associated with a G-mode psalmody, and both antiphons have a G final in the St. Peter's antiphoner (fol. 21): "Karolingische Renaissance," p. 7. 74 With respect to Italy see Paul Merkley, "The Transmission of Tonaries in Italy," Studies in Music from the Universityof WesternOntario 10 (1985), 166-224. 75 When the Old Roman repertoire was supplemented with Gregorian responsories for the feast of the Apparitio of St. Michael, the Gregorian responsory tones were used instead of the special Old Roman ones.

The Singing of Psalms 559 parallel to that which existed in the Gregorian tradition during its pretheoretical stage. An alternative hypothesis could be suggested. Did the choral singing of psalms implant itself in Rome, at least where Old Roman chant was sung, at a much later date than it did elsewhere in Italy and northern Europe? Such a late transformation, allied to the lack of a theoretical tradition in Old Roman chant, would have fostered the soloistic diversity I have postulated. Although Old Roman psalmody before the twelfth century cannot be recovered, its configuration at that point fits well with the hypothesis of a derivation from a highly varied solo psalmody not fully controlled by the pressures toward conformity exercised by the tonaries and their theorist-compilers. Perhaps also connected with the introduction of choral psalmody, if only peripherally, was the introduction to Rome of monumental choir enclosures for the chanting of the Divine Office.76 These are first documented at Rome during the pontificate of Paschal II (1099-1118), a former Cluniac monk. Their erection symbolized a higher degree of solemnization of the canonical Office and may have also signified a new and more important role for the "chorus psallentium." In my compilation of Old Roman differentiaefor the antiphonal psalmody of the Office (Appendix C) I have chosen a "modal" arrangement of the formulae. This allows comparison with other published compilations and demonstrates that in Old Roman psalmody there are elements which offered a foundation for the Gregorian system of finals and related reciting tones elaborated by Frankish theorists. Most obvious of these elements is the absolutely consistent choice of reciting tone(s) with a given final. There are but two striking features which set Old Roman psalmody apart: the absence of a c reciting tone with an F final and the very frequent appearance of an E final followed by a psalm formula which recites on the final itself.77 This latter procedure, usual with Old Roman antiphons of the Office, is comparatively rare in the Gregorian tradition. Its appearance is restricted to a very few manuscripts. I have found it in only two Gregorian antiphoners from Italy (Ivrea 62 and Cividale 57; see Ex. 6b). Dom Claire has pointed out its presence with an antiphon model he calls "timbreK" (LaudaJerusalem represents this type) in manuscripts thought to preserve archaic psalmodic practices, among them Metz 461 and Aachen, Cap. 35.78 It occurs also in antiphoners from Cambrai and Ivrea (Exx. 6a and b) as well as in the Old Roman antiphoner of St. Peter's (Ex. 6c). Antiphons with an a final and a reciting tone on the final have a similar intervallic context, though the flat is
All of the available information on these enclosures is assembled and interpreted by Elaine DeBenedictis, "The 'Schola Cantorum' in Rome during the High Middle Ages," Ph.D. dissertation, Bryn Mawr College, 1983. 77 Another anomaly is of more restricted significance: the "0" antiphons of Advent, all with D final, are associated' with a psalmody reciting on E, but with a differentia not encountered elsewhere in conjunction with E-mode antiphons: Vat., Arch. di S. Pietro B 79, fols. 14v-15r; London, Brit. Lib., Add. 29988, fol. 14r-v. For a different arrangement of the Old Roman psalm tones see Hucke, "Karolingische Renaissance," pp. 15-17. 78 "L'office ferial," pp. 99-105.

The Singing of Psalms not always expressed (Ex. 6d).79 The Beneventan tradition preserves a number of these cases attached to the antiphons Speret Israel, In matutinis, Quia mirabilia,Iubilate Deo.80 560
EXAMPLE6 Psalm tones with reciting note on final of antiphon A (a) CambraiC.38, fol. 52v Lau da ihe-ru - sa-lem do-minum. e u o u a e

Ivrea 62, fol. 55v



(thirdhigherin MS)


ie-sum chris-tum do- mi num.

Arch. di San Pietro, B79, fol. 55r

'2 ^^
Lau -

' ^


da ihe-ru- sa-lem

Turin,F.I.4, fol. 120r



ie- ru - sa- lem

do- minum.

Antiphonale Monasticum,p. 162r Lau da Je-ru - sa- lem do- mi num.

The Old Roman chant antiphoners give exclusive preference to this E-

Cividale, Museo Archeologico Nazionale 57, fol. 60v. Benevento, Bibl. cap. V.19, fols. 180v, 183r, and 187v. Thomas Kelly has informed me that he has discovered on flyleaves in a private collection (photographs at Solesmes) certain Beneventan antiphons with a D final associated with a reciting tone on E. Aurelian of R6eme recognized a psalmody on the final (D, in this case) for the antiphons Nos qui vivimus, Martyres Domini, and Angeli Domini (Musica disciplina 16.29, ed. Gushee, pp. 110-11). Michel Huglo has noted a similar practice maintained in some French churches "for many centuries" with these antiphons, familiar because of their usual association with the tonus peregrinus; see "The Tonus Peregrinus - A Question Well Put?" Orbis Musicae: Studies in Musicology (Tel Aviv, 1980), pp. 5-6. In the eighteenth century the Abbe Lebeuf observed in contemporary French antiphoners many similar cases of a reciting tone on the final of the antiphon or on the tone above the final. He considered these anomalies survivors of Gallican chant: Traite historiqueet pratiquesur le chant brevis, an acknowledged repository of ecclesiastique(Paris, 1714), pp. 32-36. The Commemoratio archaic psalmodic customs, treats such D psalmody as a special, but by no means exceptional, category (ex. 27, pp. 54-55, in the Bailey edition). Charles Atkinson surmises that the parapter tones associated in theoretical sources with the three troublesome antiphons mentioned above might be "a remnant of an earlier, more flexible, and perhaps even non-Roman, practice": "The Parapteres:Nothi or Not?" The Musical Quarterly68 (1982), 51.
79 80

561 The Singing of Psalms psalmody and have no trace of the typical Gregorian mode-III psalmody reciting on c, which is, however, the norm in the Old Roman psalmody of the Mass. Dom Claire has postulated a theory of evolution from a "modalite archaique" (psalm recitation on the final), which would require in this case that the E "corde-mere" remain the reciting tone while the final descended to A (with obligatory B-flat). Transposed up a fifth, this would result in a reciting note on b. A hypothetical subsequent development caused the supposedly "unstable" b-natural to drift towards c, the customary reciting note in Gregorian mode-III psalmody.8' This line of reasoning regards preference for the b reciting tone as evidence of an earlier stage of development. A corollary of the same hypothesis presumes to explain why the reciting tone in Gregorian tone VIII is c and not b, as analogies with modes II and VI might suggest. One finds a mixture of b and c reciting tones in Beneventan chant (Ex. 7).
EXAMPLE7 Beneventan psalm tones with reciting note on b or c




BCG in lower MS) (tone


Q Us



I .I -'

"e3{ --



* N Z ' AC

'Z H^

A = Benevento V.19-20 B = Benevento V.21 C = Benevento V.22 D = Monte Cassino 420 E = Monte Cassino 542 F = Monte Cassino, CompactioV G = Monte Cassino 318 H = Naples, Bibl. Nat. Cent., XVI.A.7 I = Vat. lat. 14446

81 Jean Claire, "L'evolution modale dans les r6pertoires liturgiques occidentales," Revue gregorienne 40 (1962), 196-211, 229-45.

The Singing of Psalms Benevento V.21 (siglum B in Ex. 7) has the strongest tradition, with a variety of formulae reciting on b and ending on the three possible cadential tones (G, a, b - indicated in the left margin of Ex. 7). Four Beneventan manuscripts (E, F, H, I) have only a single tone reciting on b, and it is the one most widely found, as can be seen in Ex. 7. None of these manuscripts is complete, however, so generalizations about their contents may be dangerous. Comparison of the contours of these psalm tones in the Beneventan manuscripts which have both b and c as reciting tones lends some support to the hypothesis of a semitone displacement of the hypothetical original b reciting tone.82 Although exact pitches can be difficult to interpret in some Beneventan manuscripts because of the frequent absence of clefs, both Benevento V.21 and 22 are quite clear in their notation and serve as a control for the other Beneventan-Cassinese sources. They both notate unambiguously the same differentia ending on G in two different versions, one reciting on b and the other on c. Such a doublet for the parallel differentia ending on a does not exist, probably because the shift up to c would make the melodic contour of the resulting differentia too static. Two of the Beneventan sources, as well as the second tonary in Monte Cassino 318, have a psalm tone reciting on b which resembles the tonus irregularis of the modern Antiphonale monasticum (see Ex. 6e above).83 A psalm-tone recitation on E may be preserved in one of the most celebrated of medieval treatises: the Musica enchiriadis. It is the musical example which illustrates composite organum at the fifth ("Sit gloria Domini").84 Some of the Enchiriadis manuscripts notate this piece incompletely or omit it altogether. Nancy Phillips has demonstrated convincingly that it is not an antiphon, but a complete psalm tone, and she assumes for it a pitch level which produces a differentia similar to the ones under discussion.85 The tone is transcribed without a clef in Ex. 8, but the starting pitch can be taken as E, b, or (with b-flat) a with the same results from the standpoint of psalmody. 562

82 The matter has been discussed by Eugene Cardine, "La corde recitative du 3e ton psalmodique dans 1'antique tradition sangallienne," Etudes gregoriennes 1 (1954), 47-52; the case for b as a primitive reciting tone is made by Joseph Gajard, "Les recitations modales des 3e et 4e modes et les manuscrits beneventains et aquitains," ibid., 9-45. The b reciting tone is the norm in the Commemoratio brevis (Bailey ed., exx. 6, 19, 52, 53, 54). Although unrelated to the Beneventan tradition, the Antiphoner of Ahrweiler preserves doublets of psalm tones in both the normal and the "Germanic" versions (see n. 26 above). 83 Monte Cassino 318, p. 259 and Ex. 6e of the present article. 84 Hans Schmid, ed., Musica et Scolica enchiriadisuna cum aliquibustractatulisadiunctis, Bayerische Akademie der Wissenschaften, Veroffentlichungen der musikhistorischen Kommission 3 (Munich, 1981), p. 42; Gerbert, Scriptores, 1:167. 85 "'Musica' et 'Scolica Enchiriadis': The Literary, Theoretical, and Musical Sources," Ph.D. dissertation, New York University, 1984, pp. 451-63. It resembles the "tonus irregularis" assigned to certain antiphons in the modern Antiphonale monasticum and an Ambrosian psalm formula reciting on E; see Terence Bailey, "Ambrosian Choral Psalmody: The Formulae," Rivista internazionaledi musica sacra 1 (1980), 316.

The Singing of Psalms

EXAMPLE8 Psalm tone reciting on E/b from the Musica Enchiriadis


Sit glo

ri - a



in sae-cu-


lae- ta- bi- tur do-mi-nus

in op - e - ri- bus su- is.

Since psalm tones are music at the service of words, one of the chief concerns of the singer (whether solo or choral) is the manner in which the psalm text is set to the cadences of the psalm tones. Though strong arguments have been made that the cursive method (i.e., no accommodation of textual accents to the melodic formula) is much older, all the evidence I have been able to uncover leaves no doubt that the adaptation of changing accent patterns is the norm from at least the tenth century. One of the chief concerns of the brevis is the adaptation of these textual patterns ("pro diversa Commemoratio positione verborum") to the cadential formulae of the psalm tones.86 The author seems more concerned about the mediants than about the finals, however. Many combinations ("modi" in his terminology) are illustrated with practical examples - "propter tardiores fratres," as he explains! The profusion of often redundant examples illustrating various patterns of textual accents would not have been needed under the older, flexible solo practice, but the exigencies of choral performance demanded clarification and simpreplification. In providing this assistance the author of the Commemoratio served many fascinating features of archaic psalmody, only some of which survive in the extant practical sources. The Italian antiphoners contain virtually no psalm tones underlaid with a complete psalm verse, which would illustrate both mediant and final cadences. Ex. 9 includes the only two specimens I have encountered; one of them is from the Old Roman antiphoner of St. Peter's. Aside from the rather elaborate mediant cadence in the example from Lucca, neither of these specimens is unusual in any respect. Most of the texts completely notated to a psalm tone are exceptional in that they are not taken from the psalter. One can easily understand why: years of daily familiarity gave the singers an intuitive mastery of the variables inherent in the 150 psalms. They did not need specific illustrations for any of the psalms, whereas texts not taken from the psalms would not be so familiar and readily adaptable. The largest category of fully notated texts is

86 For a treatment of the possibilities see Ruth Steiner, "Cursus," TNG, 5:99-101, and the extended essay in PM 4, pp. 27-204. A valuable critical examination of medieval and modern (Solesmes) practices is Terence Bailey, "Accentual and Cursive Cadences in Gregorian Psalmody" (n. 45 above).


The Singing of Psalms

Completely texted psalm tones

Lucca, Bibl. cap. 602, fol. 58r

Miserere -w me i de us

e--m-' -w w di -

am tu- am.

se- cun- dum mag-nam mi- se- ri- cor -

di Arch. S. Pietro
B 79, fol. Ilr Utconfirmet il-lud et cor-ro-bo-retin iu-di-ci-o

et iu-sti-ti- a

a - mo- do


us- que


sem- pi - ter-


the series of versus ad repetendumfor nocturns and lauds on the feasts of the

following manuscripts examined in this study have either full (nocturns and lauds) or partial sets of the versus. (For complete documentation on the manuscripts see Appendix A.) St. Paul Vercelli 37, fol. 56v Vercelli 70, fol. 116r VallicellianaC.5, fol. 223v (natale) VallicellianaC.5, fol. 79r (conversio) VallicellianaC.13, fol. 219v Vat. lat. 14676, fol. 162v St. Laurence Vercelli 37, fol. 135r VallicellianaC.5, fol. 226v

of St. Paul (January

25) and St. Laurence


10).87 The

Vat. lat. 14676, fol. 169v

87 The short texts are drawn from autobiographical writings of St. Paul and from the passio of St. Laurence. They are intended to be sung after the "Gloria patri-Sicut erat" at the end of the psalm and are followed by a final repetition of the antiphon. See Honorius of Autun, Gemma animae 4.115: "Nocturnale officium de sancto Paulo ideo versibus antiphonarum insignitur, quia ipse plus omnibus laborasse apostolis legitur. Similiter versus ad antiphonas de sancto Laurentio cantatur, quia eius passio omnibus martyribus praefertur, sic de ceteris notandum est" (PL 172:732). Amalarius of Metz does not mention them in connection with the feasts of St. Paul or St. Laurence in his Liber de ordine antiphonarii 60-61, ed. Jean Michel Hanssens, Studi e Testi 140 (Vatican City, 1950), p. 97. As far as I am aware, the tradition of these verses and the order in which the customary texts appear have not been studied. They were observed in French manuscripts by Amedee Gastoue, "La psalmodie ancienne des huit tons," La tribune de SaintGervais 14 (1908), 196.

The Singing of Psalms

Udine, Bibl. arch. 79, fol. 199v Udine, Bibl. arch. 84, fol. 133v Cividale 57, fol. 169r Gorizia B, fol. 178v Monza 16/82, fol. 141v Ivrea 62, fol. 156v Benevento V.20, fol. 213v Udine, Bibl. arch. 79, fol. 218r Udine, Bibl. arch. 84, fol. 138r Gorizia B, fol. 204v Monza 16/82, fol. 145v Ivrea 62, fol. 160r Ivrea 33, fol. 54r Benevento V.20, fol. 250r


Vallicelliana C.5 also has versus ad repetendumfor the feasts of the Assumption (fol. 230v) and St. Denis (fol. 255r), while Ivrea 62 contains special ones for the feast of St. Syrus, first bishop of Pavia (December 9, fol. 139v). The uniformity of the antiphoners in transmitting these versus suggests that the texts were not adapted spontaneously by the scribes: they were copied from the model just as the surrounding antiphons were. Nevertheless, they are the largest body of material for judging how prose texts were set to the psalm tones. In all cases the differentiaeare adapted to the variable accent patterns of the text. Since the manuscript antiphoners transmit only the concluding cadence of the psalm tone, mediant cadences go unrecorded. The examples in the Commemoratio brevisand the contents of the antiphoners listed above represent the largest body of medieval material on the internal cadences of the psalm tones. Not all modes are found, because not all modes are represented among the antiphons to which the written-out versus ad repetendumare attached. A few of the tones have mediants which are as elaborate as some final cadences (Ex. 10).

of Mediant cadences versus repetendum ad (a) Vercelli37, fol. 136v

Gra- ti as ti - bi a go do mi- ne...

De - us F.IV.4, (b)Turin,
fol. 170v

pre-or-di- na-vit me ut vi-de - rem iu - stum...

Udine84, fol. 139r


Gra-ti -

as ti- i a- go do-mi-ne ihe as ti - bi a- gs do-om

su chri - ste... ste...

Lau- da - te

do- mi- num om- nes gen -


The Singing of Psalms Sometimes the texts are so brief that no mediant cadence occurs. All of the written-out tones that I have seen in the antiphoners must be interpreted as accentual cadences.88 None of them are treated cursively. This is true of both the mediant and the final cadences, just as it is in the Commemoratio brevis. Only three modes (I, VII, and VIII) offer sufficient writtenout examples in the Italian antiphoners to establish treatment of accent at the final cadence, but all of them are accentual. Modes I (Ivrea) and VIII (Vercelli) adjust to the final accent only, but mode VII (Vercelli) adjusts to the last two accents of the line. Ex. 11 demonstrates accentual mediant cadences and reflects a general characteristic of mode-II mediants found also in the Commemoratio: elevation of the reciting pitch a few syllables before the the final accent. 566
I EXAMPLE of Elevation reciting tonebefore mediant cadence

Cum in- tras- set ihe - sus in De - us qui ha - bi Qui - a ip - se

ternmplum ta - re do - mi - nus

de - i... fa- cis... no- vit...

Benevento, V.21, fol. 99r Benevento, V.21, fol. 129v Benevento, V.20, fol. 251r



In re- ge- ne-ra- ti - o- necum se- de-ret fi - li- us ho-mi-nis... Benevento, V.20, fol. 214v

A() n
Et erunt ut complaceant Caeli enarrant e - lo- qui- a

me-i... o - ris glo- ri- am de- i... Comm.Brevis, Bailey ed., ex. 40 Comm.Brevis, Bailey ed., ex. 41

Final cadences are treated similarly. Ex. 12a illustrates this with passages from the versus for St. Paul and St. Laurence which show adaptation of the formula for a line of text which ends with a proparoxytone. Ex. 12b shows a differentia adapted to accommodate a proparoxytone; by chance the same differentia appears in its "normal" state elsewhere in the Bobbio antiphoner (Turin, F.III.8). The initia of the psalm tones, which connect the final note of the antiphon with the reciting pitch, are generally those found in the modern chant books. The written-out medieval examples show widespread preference

Given this unanimity, it is extraordinary that the Instituta patrum (early 13th c.) gives special emphasis to the cursive cadence. S. J. P. van Dijk, "Saint Bernard and the Instituta Patrum of Saint Gall," Musica disciplina 4 (1950), 99-109.

The Singing of Psalms




Adaptationof final cadence to differing accent patterns (a)^ (a) Ivrea 62, fol. 156v fol. 157v fol. 156v Vercelli 37 fol. 135r fol. 135v fol. 57v (b)^ (b) Turin,F.III.8, fol. 131v ...gra ro e -go ti - am iu na
E-i & I

mun su sti

do. -am. ti - e.



L. /.

k @-pa ter -

ni ...gra ... chri -

ta ti stum

'* (k
ti am con tu

de fess - us

a. i. sum.

I.. .

L. /.



. '^^
e u o u a e

I '

- .,. .


hoc nunc et us-que in se- cu-lum.

Ivrea 62, fol. 140v e u o u a e [in- ter-ces-] si - o- ne cor-ro-bo ra.

for a "second intonation" to introduce the last half of the psalm verse (Ex. 13).

EXAMPLE13 Psalm tones with second intonation Vallic. C.5, fol. 255v Et e - ter - nita - tis sub - stan- ti - a un- um

et pr- s- ns'
et per- so-nis Monza 15/79, fol. 127v





tri-num de-um con-fi-

ten -

~~~~~e. ~~~~~~~~~~~~~~~M
Be- ne- dic a - ni- ma me - a do- mi- no, et om- ni- a...

The Commemoratio tacitly assumes that this second intonation is a normal feature of psalm tones, a view which finds confirmation in the Italian antiphoners. The modern chant books do not reflect this medieval tradition.

I would now like to review some of the insights gained from this study of psalm singing in manuscripts of the medieval Office. Textual documentary

The Singing of Psalms evidence previous to the eighth century points to solo rendition of the psalms in the monastic Office. At that period the responsibility of singing the psalms was shared by the entire community in succession, and hence required no unusual musical ability. This practice presumably led to a considerable diversity in the formulae used to chant the psalms, particularly since there existed no theoretical restrictions to the introduction of variant formulae. The advent of choral psalmody stimulated theorists to standardize the melodies to which the psalms were sung and to prescribe the class of antiphons with which each formula should properly be used. Choral psalmody also entailed a progressive campaign to reduce the diversity inherited from an earlier age, a process which naturally found an echo in the practical sources. Though the extant sources do not present a compelling pattern of inexorable reduction, the large repertoire found occasionally in the twelfth century disappears completely by the fourteenth. The sole eleventh-century source, the noted breviary Todi 170, still maintains a fairly large repertoire of differentiae (40). The aim of the differentia system was to permit all the monks to participate simultaneously in the sung psalter. At the time this system originated, the antiphon must still have been intercalated with some frequency; otherwise the differentiae would have been superfluous. Centuries later, after the antiphon had been reduced to a simple frame for the singing of the psalm, the entire differentia apparatus seemed, and indeed was, both cumbersome and obsolete. A few aspects of an earlier age are also recoverable in Italian and northern antiphoners from the twelfth century: a generally more varied differentia system, psalm tones reciting on the final (E or a; most prevalent in Old Roman chant), and absence of the equivalent of Gregorian mode V in the chants of the Old Roman Office. The reduction in psalm tones left a core of common differentiae found nearly everywhere, but there persisted a few regionally preferred formulae. The b reciting tone, found in Beneventan manuscripts, is probably just such a regional and perhaps archaic feature. All the evidence in the practical sources points to adaptation of the psalm tone to the changing accent patterns of the text.89 The system of psalmody revealed by the practical sources does not differ essentially from that recorded in the theoretical tradition, except for evidence of a diversity of formulae prevalent before the earliest antiphoners with notation that can be accurately transcribed. This diversity might have lasted even longer, were it not for the replacement of solo chanting of the psalm verses by choral psalmody in the late eighth century. This latter development, as much a part of the history of spirituality as of the history of music, occasioned a breach in the musical traditions of western monastic psalmody.90 The force of tradition was powerful enough, however, to preserve a few elements which link the medieval Office with its roots in the most ancient psalmodic traditions of cenobitic monasticism.
89 As we have seen, this procedure is assumed to be the prevailing one by the Commemoratio brevis. 90For a brief discussion see Dyer, "Monastic Psalmody of the Middle Ages."


The Singing of Psalms

APPENDIX A LIST OF MANUSCRIPTS Noted Psalters Manuscript Bibl. Vat., Chigi C.VI.177 Bib. Vat., Chigi A.VI.163 Udine, Bibl. arch. 72 Bibl. Vat., Arch. di S. Pietro E 14 Vercelli, Bibl. cap. 66 Date 11/12 c. 12 c. 13 c. 13 c. 14 c. Differentiae (few) (few) (few) (few) (few)


Comments MLBV no. 149; fragmentary; gaps in notation MLBV no. 13; some notation added in 13-14 c. fragment; notation not always entered music on fols. 31r-150r; pitches not always clear incomplete

Antiphoners Manuscript Benevento, Bibl. cap. V.21 Florence, Arch. del Duomo Florence, Bibl. Laur., Conv. sopp. 560 Ivrea, Bibl. cap. 62 (olim 64) Klosterneuburg, 1010, 1012, 1013 London, Brit. Lib., Add. 17302 London, Brit. Lib., Add. 29988 Lucca, Bibl. cap. 599 Lucca, Bibl. cap. 601 Lucca, Bibl. cap. 603 Monte Cassino, Compact. V Monte Cassino, 542 Monza, Bibl. cap. 15/ 79 Piacenza, Bibl. cap. 65 Rome, Bibl. Vallic. C.5 Date 12 c. 12 c. 12 c. Differentiae 45 33 24 Comments from S. Lupo (?); 304 fols.; CAO 5, no. 615 from cathedral; CAO 5, no. 240; incomplete 228 fols.; lacunae

12 c. 12 c. 12 c. 12 c. 12 c. 12 c. 12 c.

45 30 26 83 33 39 40

260 fols. Germanic dialect; CAO 5, no. 267 123 fols. Old Roman antiphoner from the Lateran (?) 360 fols. PM 9; CAO 5, no. 709 257 fols.; many gaps in notation; from S. Maria of Pontetetto 113 fragments; pitches not easily determined incomplete (194 pp.); psalmody related to above 275 fols. antiphoner: fols. 274r-450r from S. Eutizio; partial inventory: Ledwon

12 c. 12 c. 12 c. 12 c. 12 c.

25 32 32 31 31

570, Udine, Bibl. arch. 84 Bibl. Vat., Arch. di S. Pietro B 79 Cambrai, Bibl. munic. C.38 Lucca, Bibl. archiv. 5 Lucca, Bibl. cap. 602 Monza, Bibl. cap. 16/ 82 Turin, Bibl. Naz. Univ. F.III.8 Turin, Bibl. Naz. Univ. F.IV.4 Bibl. Vat., Vat. lat. 14676 Bibl. Vat., Borg. lat. 405 Vercelli, Bibl. cap. 70 Cividale, Mus. Archeol. Naz. 57 Gorizia, Bibl. del Semin. Teologico A Gorizia, Bibl. del Semin. Teologico B Turin, Bibl. Naz. Univ. F.I.4 Udine, Arch. cap. 30 and 26 Bibl. Vat., Arch. di S. Pietro B 87 Vercelli, Bibl. cap. 37

The Singing of Psalms

12 c. 12 c. 13 c. 12/13 c. 12/13 c. 12/13 c. 13 c. 12/13 c 12/13 c. 13 c. 13 c. 14 c. 13/14 c. 67 58 35 32 42 34 24 30 50 32 31 24 28 204 fols.; from Treviso MLBV 118; Old Roman antiphoner pp. 1-394 225 fols. CAO 5, no. 329; 236 fols. from Bobbio; 233 fols. from Bobbio; 287 fols.; faded, damaged at top of pages MLBV 140; from Pavia; 244 fols. MLBV 456; 226 fols. 288 fols. Germanic dialect; CAO 5, no. 194 Germanic dialect; from Aquileia (?) 344 fols.; calendar of Aquileia; Germanic chant dialect 335 fols; from Bobbio 189 fols. and 94 fols.; with Arch. cap. 24 and 20 part of a 4-vol. antiphoner 329 fols. 182 fols.

13/14 c.


14 c. 14 c.

36 30 (ca.) 23 28

14 c. 14 c.

Noted Breviaries Manuscript Monte Cassino 420 Todi, Bibl. com. 170 Rome, Bibl. Vallic. C.13 Benevento, Bibl. cap. V. 19-20 Benevento, Bibl. cap. V.22 Date 11 c. 11 c. 12 c. Differentiae 23 40 31 Comments 422 pp.; few diff.; heighted neumes; ODMA 53-54 544 pp.; ODMA 70-71 403 fols.; (Adv.-Easter); from S. Eutizio; inventory: Ledwon

12 c. 12 c.

50 (ca.) 54

CAO 5, no. 137 216 fols.; from Benevento

The Singing of Psalms

Naples, Bibl. Naz. Centr. XVI.A.7 Rome, Bibl. Casanat. 1574 Bibl. Vat., Vat. lat. 14446 Bibl. Vat., Chigi C.V. 137 Vercelli, Bibl. cap. 170


12 c. 12 c. 12/13 c. 13 c. 13 c.

26 25

19 24

from S. Deodato (?); many lacunae 375 fols.; from Gaeta; partially noted; ODMA 65-66 a fragment (63 fols.); MLBV 487; from Caiazzo MLBV 242; 147 fols. (summer only) from S. Pietro de Castro (?)

CAO Ledwon

Rene-JeanHesbert, ed. Corpus Officii.6 vols. Rerum Ecclesiasticarum antiphonalium Documenta:Series Maior,Fontes 7-12. Rome, 1963-79. Jacob Ledwon, "The Winter Office of Sant'Euteziodi Norcia,"Ph.D. dissertation, State Universityof New York at Buffalo, 1986.
Pierre Salmon, Les manuscrits liturgiques de la BibliothequeVaticane, 1 (Vatican City,


Paleographie musicale.

1968). divinau moyen Lex orandi 43. Paris, 1967. Pierre Salmon, L'Office age.




Aur 5 1 4 5 1 1 10 5 x 33

Reg 5 1 5 5 2 1 6 3 28

MC-1 13 3 3 11 4 3 9 8 54

MC-2 10 1 5 9 2 1 7 6 41

Cas 9 1 4 7 1 1 6 4 33

Har 9 2 6 8 2 2 6 6 41

Clm 9 2 6 9 2 2 6 5 41

Pia 8 1 4 6 1 1 8 5 34

11 2 7 10 3 2 13 7 55

ABBREVIATIONS: Metz Aur Reg MC-1 MC-2 Cas Har Clm Pia Flo Ver Mon LU

Walter Lipphardt, Der karolingischeTonar von Metz. Liturgiewissenschaftliche Quellen und Musica disciplina, ed. Laurence Gushee. Corpus scriptorum de musica 21. Rome, 1975. de Tonary of Regino, ed. Edmond de Coussemaker, Scriptorum musica medii aevi nova seri 73. Monte Cassino, Abbazia, Q 318 (11 c.), pp. 128-56. Monte Cassino, Abbazia, Q 318 (11 c.), pp. 245-85. Rome, Biblioteca Casanatense 54 (11 c.), fols. 102v-103r. Marginal letters added to the Antiphoner of Hartker (12 c.), St. Gall, Stiftsbibliothek Paleographie musicale, ser. 2, vol. 1, p. 50*. Munich, Bayerische Staatsbibliothek, clm 14523 (12 c.), from information in Michel Hu Piacenza, Archivio capitolare 54 (12 c.), fols. lr-7r. Florence, Archivio del Duomo (12 c.), fols. 277r-283r. Vercelli, Biblioteca capitolare 70 (13 c.), fols. 213v-222r. Monza, Biblioteca capitolare 16/82 (13 c.), fols. 218v-224v. Liber Usualis (Tournai, 1956), pp. 112-17. Antiphonale monasticum(Tournai, 1934), pp. 4*-30*.

The Singing of Psalms



D-Mode: Reciting tone on a


e u o u a


e u o u a


B 79 only


BL only



mm 0 -


BL only


BL: aG

,r only BL

|B79: aGFE


r~~~~~~~~~~~~~~~ ~I ~~~~~~I1-$

m --


BL only





BL only _

V BL only &^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~I




w i

D-Mode: Reciting tone on F

, m, mA aN . 1 a B79only e u o u BL: E I e A I8V n



- - j-


e u o u a

W.O I .

BL only

h- .


Used with "0" antiphons only

sed -+


s on


The Singing of Psalms E-Mode: Reciting tone on E


a I e B 790nly e -, o u a u


e u o u'a


'V BLonly



B 79 only B 79 only L


BL only


E-Mode: Reciting tone on a

BL only

e u o u a

BLonly BLonly
M9 only

euou e u o u a


BL only


B o790nly

BL only

V BL only
B 79: GF | A




B 79 only

BL only


BL only

The Singing of Psalms F-Mode: Reciting tone on a



e u o u

a I

e I

B 79only

e u o u


79: GP

B 79 only

B 79 only

L-' on1

* =^ * *

I* IT,

BL only

BL only

BL only

V BL only


BL only


The Singing of Psalms

G-Mode: Reciting tone on c
~A ~
BL: ba k

" e uu o u
BL/B79: aG


BLonly e u o u ,a I e
BL/B 79: a I


BL only

BL/B 79: c |

BL only

BL: cdG




_m !m

mI 1 ,

BL only


I ~~~~~~ | ~BL: cbaG

BL: a BL only [

BL only

BL only

The Singing of Psalms

G-Mode: Reciting tone on d



u a u

B 79 only e u o u

a I e

V BLonly

V B 79 only

B 79 only


BL:cb I BL: cb I



s A


BL: c I

A AE @



BL: dc

)V B 79 only

R 79: d



a-Mode a-Mode

V B79only

e u o u

V BLonly

e u o u

B 79 only


The Singing of Psalms


The Old Roman antiphoners exhibit a remarkable degree of diversity as well as disagreement in their repertoire of psalm tones. Although the total number of differentiae surpasses 100, many of them are confined to only one of the antiphoners. That has been indicated in Appendix C. In the absence of an indication that a particular formula is found only in the St. Peter's manuscript (B 79) or the antiphoner in the British Library (BL), it may be assumed that it is common to both. About one-third of the repertoire falls into this category. The greatest agreement occurs in D-mode psalmody with an a reciting tone; the greatest disagreement occurs in F-mode psalmody with an a reciting tone. Curiously, the Gregorian equivalent of the latter (mode VI) shows the greatest agreement. The E-mode psalmody with an a reciting mode is in one respect another nexus of disagreement between the Old Roman antiphoners. In a few cases I have indicated variant readings above the staves. Further consideration of the material could result in their classification as independent differentiae. I have tried to be conservative in my estimates, and I have not included in Appendix C the psalmody of the Paschal Vespers found in the Old Roman graduale, Vat. lat. 5319.

Joseph Dyer is Associate Professor of Music at the University of Massachusetts,Boston, MA 02125.